P. 1
Project Planning, And Control

Project Planning, And Control

|Views: 618|Likes:
Published by Saak Junior

More info:

Published by: Saak Junior on Jun 06, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

06/04/2014

pdf

text

original

Project Planning, and Control

Also available from Taylor & Francis
Construction Project Management
Peter Fewings and Martyn Jones
Taylor & Francis Hb: 0–415–35905–8
Pb: 0–415–35906–6
Construction Safety Management Systems
Steve Rowlinson
Spon Press Hb: 0–415–30063–0
Human Resource Management in
Construction Projects
Martin Loosemore, Andrew Dainty and Helen Lingard
Spon Press Hb: 0–415–26163–5
Pb: 0–415–26164–3
Occupational Health and Safety in Construction
Project Management
Helen Lingard and Steve Rowlinson
Spon Press Hb: 0–419–26210–5
The Management of Construction Safety
and Health
R. J. Coble, T. C. Haupt and J. Hinze
Spon Press 90–5809–328–X
Information and ordering details
For price availability and ordering visit our website www.tandf.co.uk/
builtenvironment.com
Alternatively our books are available from all good bookshops.
Project Planning, and
Control
David G. Carmichael
First published 2006
by Taylor & Francis
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
Simultaneously published in the USA and Canada
by Taylor & Francis
270 Madison Ave, New York, NY 10016, USA
Taylor & Francis is an imprint of the Taylor & Francis Group
© 2006 David G. Carmichael
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical, or
other means, now known or hereafter invented, including photocopying
and recording, or in any information storage or retrieval system,
without permission in writing from the publishers.
The publisher makes no representation, express or implied, with regard
to the accuracy of the information contained in this book and cannot
accept any legal responsibility or liability for any efforts or
omissions that may be made.
British Library Cataloguing in Publication Data
A catalogue record for this book is available
from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book has been applied for
ISBN 13: 9–78–0–415–34726–6
ISBN 10: 0–415–34726–2
This edition published in the Taylor & Francis e- Library, 2006.
“To purchase your own copy of this or any of Taylor & Francis or R outledge’s
collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”
(Print Edition)
Contents
About the author vii
Preface viii
Notation x
PART I
Conventional treatment from a systems
perspective 1
1 Introduction 3
Background 3
Replanning 12
Planning effort 21
Book outline 24
2 Systems thinking 29
Introduction 29
Systems–subsystems 29
Single level system 32
Fundamental systems problems 36
3 Planning tools and terminology A–Z 40
4 The planning process 122
Introduction 122
Planning problem components 124
Simplifications to planning 127
Objectives and constraints 129
Scope 134
vi Contents
Extras feeding into and out of planning 137
Programs 138
5 Planning over levels 148
Introduction 148
Hierarchical modelling 149
Work breakdown 152
Iterative analysis approach 166
Example 168
6 Replanning 179
Introduction 179
Monitoring 186
Popular approaches to reporting 190
Replanning representation 209
7 Selected topics 220
People issues 220
Monte Carlo simulation 228
Linear projects 249
PART II
Synthesis treatment 267
8 Multistage planning 269
Introduction 269
Stage modelling 269
Optimisation problem 275
Examples 276
9 Synthesis over levels 284
Direct synthesis approach 284
Models 284
Level planning problems 292
Multilevel optimisation thinking 295
10 Selected topic 305
Project compression 305
Index 324
About the author
David is a graduate of The University of Sydney (BE, MEngSc) and The
University of Canterbury (PhD) and is a Fellow of The Institution of Engi-
neers, Australia, a Member of the American Society of Civil Engineers,
formerly a Graded Arbitrator with The Institute of Arbitrators, Australia,
and a trained mediator. He is former Head of the Department of Engineer-
ing Construction and Management and currently a Consulting Engineer
and Professor of Civil Engineering at The University of New South Wales.
He has acted as a consultant, teacher and researcher in a wide range
of engineering and management fields, with current strong interests in all
phases of project management, construction management and dispute reso-
lution. Major consultancies have included the structural design and analysis
of civil and building structures; the planning and programming of engineer-
ing projects; the administration and control/replanning of civil engineering
projects and contracts; and various construction and building-related work.
In addition there have been numerous smaller consultancies in the struc-
tural, construction and building fields. He has provided expert reports and
expert witness in cases involving structural failures, construction accidents
and safety, and contractual and liability matters.
He is the author and editor of seventeen books and over sixty-five papers
in structural and construction engineering and construction and project
management.
Preface
It is interesting to watch the development of a discipline over the years.
There have been many contributors to project planning. The author was for-
tunate enough to learn the fundamentals fromJames Antill (see for example,
J. M. Antill and R. W. Woodhead, Critical Path Methods in Construction
Practice, John Wiley and Sons, New York, 1st ed. 1965). When Construc-
tion Engineering Networks (D. G. Carmichael, Ellis Horwood/John Wiley
and Sons, Chichester, 1989) was written, it seemed a losing battle to con-
vince practitioners to use network methods for planning. Some practitioners
were convinced, but the majority of planners saw network methods in a
negative light. And that was at a time when reasonably user-friendly com-
puter packages were available. Today such packages are universally used, if
only as a way of generating bar charts. The challenge now is to understand
what planning actually is, and this is the basis of this book; people are going
through the motions of planning but do not understand the fundamentals
of their trade. As such, the book treads new ground.
A systems approach is adopted, with control systems terminology pre-
ferred for its precision and usefulness over conventional planning terminol-
ogy. In particular, state space and related terminology of modern control
theory is preferred.
The bookassumes that the reader is familiar withelementary planning, as is
undertakenonall projects onadailybasis. If usedas anintroductorytextbook,
supplementary material such as given in Carmichael (1989) would be useful.
A main thrust of the book is seen as sending planning thought in the
correct direction.
Project management is now commonplace in most walks of life ranging
from the business world, service industries to technological endeavours.
Central to project management is planning. Most matters in managing
projects, whether it relates to time, costs, resources and so on, involve
planning. Planning is a crucial issue in project management and one on
which many other aspects develop.
Existing treatments on planning jump head first into network analysis,
both deterministic (CPM – critical path method) and probabilistic (for
example, PERT – program evaluation and review technique, and Monte
Preface ix
Carlo simulation), with all refinements including overlapping relationships,
and the use of industry-preferred packaged software. The author, amongst
many, is guilty of such an uninspired and pedestrian approach (Carmichael,
1989), justified falsely by the reason of ‘practicality’.
A more inspired approach is possible. In fact, planning can be shown
to be a systems synthesis or inverse problem. As such, the solutions to
planning problems are not unique. In most cases, planners are only after a
satisfactory solution, or a solution that they can live with, and do not spend
the additional time searching for the optimal solution. A planner may also
be under time pressures to come up with quick solutions.
However, planners expediently reverse the logic and deflect attention
from their inability to come up with best solutions, on-time pressures and
pseudo ‘practicality’ arguments, when in fact planners do not understand
the synthetic nature of their job. Planners are unaware of and do not
understand the components of the synthesis problem, and so they never
know where they are relative to the optimum.
The book presents a totally new approach to project planning, and relates
this to existing treatments. The aims of the book are to understand project
planning and to contribute to thinking on project planning. One of the over-
riding reasons for writing this book is to counter the myriad of misconcep-
tions and thinking errors that exist among writers on planning. The level of
thinking that goes into planning in many cases is very superficial and cook-
book in nature. To counter this, the book adopts a systems view to provide
a rigorous framework. Rigour in the usage of terminology is also stressed.
The book aims to develop the reader’s professional skills and thinking in
the planning component of project work, to enable the reader to perform
more effectively, to understand project planning procedures and to gain
an insight into the associated skills. Numerous case examples from diverse
industries, and exercises support the approach.
The book is aimed at lecturers and students (postgraduate and under-
graduate) in project management, and practitioners involved in projects (all
industries).
It contains case examples and exercises and hence could be used as a
textbook.
Organisations, for which the material is applicable, include universities,
professional organisations in engineering (all types), architecture, building
and project management.
Nothing equivalent exists. Currently most people use project manage-
ment texts that have planning (and ‘control’) as one component. The book
rethinks planning practices, and outdates thinking in existing texts.
Acknowledgment
The book contains a number of case examples contributed by many people.
Their contribution is gratefully acknowledged.
Notation
Continuous time version
i time
i
S
project start time
i
F
project finish time
i
now
present time
u(ii control at time i
x(ii state at time i
z(ii output at time i
i(ii disturbance at time i
Discrete time/staged version
l period/stage counter, l =0, 1, 2, . . . , N−1
N number of periods/stages
x(li planned-for state after period/stage l−1
x(0i initial state
x(Ni final or terminal state
xA(li actual state after period/stage l−1
x

(li revised planned-for state after period/stage l−1
var(li variance; difference between planned-for state and actual state
after period/stage l−1
u(li control at period/stage l
i(li disturbance; anything that prevents x(l+1i being
obtained exactly
Activity information
TF Total float
FF Free float
IF Interfering float
EST Earliest start time
EFT Earliest finish time
Notation xi
LST Latest start time
LFT Latest finish time
LT Lead time period
SOS Sum of squares
Part I
Conventional treatment
from a systems perspective
Chapter 1
Introduction
Background
Many people are aware of the humorous paradoxical sign:
It also reflects a common saying, but can you plan any way other
than ahead? In a like vein, people also refer in a redundant fashion to
‘future plans’, or if a project is not going well, it is said to ‘lag behind’
schedule.
Popular adages of planning
There are a number of adages and popular quotations involving planning:
In all matters, success depends on preparation. Without preparation,
there will always be failure. (Confucius)
Action without planning is the cause of every failure.
Mediocrity is so easy to achieve, there is no point in planning for it.
If you fail to plan, you plan to fail.
4 Conventional treatment from a systems perspective
If you don’t know where you are going, you will end up somewhere
else. (L. Peter)
If you don’t know where you are going, any road will take you
there.
The nicest thing about not planning is that failure comes as a com-
plete surprise, rather than being preceded by a period of worry and
depression.
4Ps Poor planning = poor performance.
6Ps Prior preparation and planning prevents poor performance.
6Ps Prior preparation prevents poor performance – probably.
Even if you do nothing, it should be done on time.
Plan the work, and work the plan.
If you plan to fail and you do, then you have been successful.
Real life is what happens when you make other plans.
The stability of a plan tends to vary inversely with its extension.
Planning also is mentioned in children’s fables. For example, one of Aesop’s
fables gives the story of the ant and the grasshopper, where the ant is indus-
trious collecting food over summer while the grasshopper plays, only for
winter to come and the grasshopper goes hungry, leading to the conclusion
that ‘it is best to prepare today for the needs of tomorrow.’ Aesop’s fable
of the cat and the fox, where the cat knows one trick to escape the dogs
and uses it, while the fox knows many tricks but cannot decide which to
use and gets caught by the dogs, concludes that ‘one good plan that works
is better than a many doubtful ones.’
Two noticeable things emerge from these adages and fables, and other
publications on planning:

Everyone says that there is a need for planning, and a need for genuine
effort to be put into planning.

Everybody has a different idea of what planning is.
Therein lies the source of most of the troubles preventing the advance-
ment of the understanding of planning. To discuss these in a project con-
text, it is first necessary to take a few steps backwards, and clarify some
thinking. The first step relates to the distinction between a project and its
end-product.
Introduction 5
Project versus end-product
A project will commonly come about because of an identified need or want
for some product, facility, asset, service and so on. This end-product is
achieved through a project (Figure 1.1).
This distinction sometimes causes people confusion and many people are
not aware of the distinction, or of the need to make a distinction. For
example, people sometimes refer to a building as a project. It is not. The
processes that go together to materialise the building are the project. The
building is the end-product. (It is acknowledged that the definition of a
project is sufficiently flexible to include the operation and maintenance
phase of a product within what is called the project. However, this is not
the issue here.)
There will be objectives and constraints relating to the end-product, and
objectives and constraints relating to the project. (The plural terms objec-
tives and constraints are generally used in this book even though the singular
forms may be applicable in some situations; the exception is where the sin-
gular is deliberately intended.) The former influences what the end-product
will look like and how it functions. The latter influences the way of getting
to the end-product. The term ‘objective’ here is used in a correct systems
synthesis sense, not as used loosely by lay people (Carmichael, 2004).
The objectives and constraints are derived from the value system of the
project/end-product owner. They lead to establishing the scope, and plan-
ning practices.
Later chapters develop a hierarchy of objectives and constraints – at
project, activity, element and constituent levels. A project is composed of
activities. An activity is composed of elements. Elements are composed of
constituents (Figure 1.2).
Planning and design
Following the distinction between an end-product and the means to the end-
product, design and planning functions are seen as parallel and fundamental
undertakings on projects. Broadly on a project, Figure 1.3 applies.
Concept
Development
Implementation
Termination
Project
Idea for
end-product
End-product
exists
End of life
for end-product
Operation, maintenance
Time
Figure 1.1 Project and end-product timeline.
6 Conventional treatment from a systems perspective
Project
level
Activity
level
Element
level
Constituent
level
Resource behaviour
Time
Figure 1.2 Hierarchical levels in a project.
End-product Means to end-product
Design Planning
Interaction
Figure 1.3 Fundamental project undertakings.
Both the planning and design problems are shown later to be inverse (syn-
thesis) systems problems, although popular practice is to solve these in an
iterative analysis fashion.
It is interesting to note that in nearly every project management text,
there is a section on planning, and invariably the text authors feel obliged
to convince the reader that there is a need for planning and of the benefits
of planning, and to justify the pages allocated to planning. The same holds
for courses given on planning. This indicates the immaturity of project
management as a discipline and as practised. By comparison, design texts
and design courses do not have to convince readers and practitioners of
the need for design; it is accepted as fundamental in order to get to the
end-product. Why then is not the parallel project task of planning accepted
as fundamental, and let the discussion get straight on to the best way to
do it?
Also of interest to note is that no owner would let a person, who has no
design education, design something for them. Owners are very discerning
Introduction 7
when it comes to design. Yet anyone can get away with calling him/herself a
planner without any planning credentials, and undertake planning for that
same ‘discerning’ owner. Yet design and planning are both fundamental
project tasks on which the project outcome equally depends. It would be
nice to think the reason for this was that the owner has to live with the visual
impact and functionality of the end-product, but can soon forget about the
cost and duration impact of getting to the end-product. But the real reason
appears to be the different maturity levels of design and planning, or more
appropriately the immaturity level of planning compared to the maturity of
design. It may also be that planning is an order of magnitude more difficult
than design, yet it is practised by people with specialist knowledge in the
inverse ratio. People have incompatible standards and expectations when it
comes to design and planning work.
Another form of undesirable imbalance can be observed in some project
managers and in some project management publications which take the
view that activity scheduling is project management; if you have a computer
package that does scheduling, then you have all that is needed to manage a
project.
The false belief in the need for planning
Everyone says that there is a need for planning, and a need for genuine
effort to be put into planning; the latter is acceptable (it is not unreason-
able to expect that effort put into planning improves the chance of project
success) but the former implies that no one recognises planning as a fun-
damental project task. Since planning is a fundamental project task, talk
of need for planning is off the mark, as is the question ‘why plan?’ posed
in books. Planning is carried out (to varying degrees of sophistication) on
every project whether people realise it or not. Without planning, a project
doesn’t proceed. Perhaps when people talk of a need for planning, they are
talking about structured planning, as opposed to ad hoc planning, rather
than talking of the no-planning versus some-planning scenario. But it is
more likely that people don’t understand the place of planning.
Whether people are involved with projects or not, everyone carries out
planning in some form covering the tasks that go to make up their daily
lives. This applies to their home roles, social pastimes, commuting between
home and work, and work tasks. Having breakfast and getting dressed
in the morning, driving from A to B, spending a weekly pay cheque etc.
are all planned to some degree. The planning might be casual or very
detailed; it might be well thought out or poorly thought out; it might be over
short or long time horizons. Without planning, the task or project does not
proceed. Different planning leads to the task or project being carried out
differently.
8 Conventional treatment from a systems perspective
What is planning?
Everybody has a different idea of what planning is. See the above adages,
for example. Project management texts give a variety of definitions. And
if lay persons, practitioners and texts on ‘strategic planning’ and ‘market
planning’ are consulted, it appears that the word planning is used by most
people in the sense of Alice in Wonderland:
‘When I use a word,’ Humpty Dumpty said in a rather scornful tone,
‘it means just what I choose it to mean, neither more nor less’.
‘The question is’, said Alice, ‘whether you can make words mean so
many different things’.
(Lewis Carroll, ‘Through the Looking Glass’, Ch. 6)
Some people refer to planning as ‘time management’, and vice versa, but
this is too restrictive, and not correct. Strictly, time cannot be managed.
In broad terms:
Planning concerns itself with the ‘means to end-product’ problem. Plan-
ning establishes how and what work will be carried out, in what order and
when and with what resources (type, and number or quantity) (additionally
expressed in a money unit).
The parallel matter of design concerns itself with the ‘end-product’
problem.
(This last sentence refers to the design function. Of course, the actual
activities involved in the design process can be treated as a subproject in its
own right and planning applied to these activities. Similarly, for those who
like embedded problems, the activities involved in the planning process can
be treated as another subproject in its own right and planning applied to
these activities. This sort of logic can be continued ad infinitum, for those
so inclined.)
‘How’ implies ‘method’. ‘Order’ implies ‘sequence’. ‘Resources’ are peo-
ple and equipment. Resource type, number (quantity) and production are
associated with method, and order and timing (start, duration and finish) of
activities. Resources (type, and number or quantity) are typically expressed
in a common measurement unit of money (Figure 1.4).
Planning outcomes provide information, for example, on:

Schedule matters (project completion date, milestones, differentia-
tion between important (critical, Pareto, . . . ) and not important
activities, . . . )

Resource matters (type, number or quantity, and timing requirements,
including outsourcing)

Production matters (related to schedule)

Money matters (project cost, cash flow, . . . ).
Introduction 9
Resource usage
Money
Figure 1.4 Some possible relationships between resource usage and money.
All tend to be strongly related. Oscar Wilde observed in Phrases and Philoso-
phies for the Use of the Young that, Time is waste of money.
Prophecy
Planning is distinguished here from prophecy.
It is essential for thinking man to believe in free will, to believe that
his future can be changed by thought and action. Prophecy denies this
and declares that all futures are immutable and fixed, that nothing
man strives towards is relevant, as is already preordained by whatever
governs his future, be it God or Destiny.
. . . But on the other hand, the answer may lie in Dunne’s concept of
alternative variable futures which run parallel to each other. Each man
receives the future dependent upon his earlier actions.
(Extracted from E. Cheetham, The Prophecies of
Nostradamus, Corgi, London, 1975, pp. 14–15)
The planning function
In systems terms, planning establishes the value of control (input, action,
decision), u(ii, throughout the project duration, i
S
≤i ≤i
F
. u(ii has multiple
parts (that is, it is a vector quantity of control variables) that contain
information on method (including sequence), resources (type, and number
or quantity and hence money) and resource production rates (or equivalent).
i is time. The project is viewed as a dynamic system.
u(ii = {method(ii, resources(ii, resource production rates(ii¦
10 Conventional treatment from a systems perspective
A hierarchical and staged approach to planning might be adopted. (Four
main hierarchical levels are considered in this book – project, activity,
element and constituent – Figure 1.2.) Decoupling the hierarchical levels
and stages facilitates the solution, but introduces iterations. Determinism
would generally also be preferred by planners.
In general, there are many possible solutions (values of control variables)
to the planning problem. The associated objectives enable a choice between
alternative solutions to give that solution which is optimal or efficient in
resource usage, most economical or best in some other way, for exam-
ple minimum duration. That is, the control is chosen so as to extremise
(optimise/minimise/maximise) the objectives while satisfying any constraints
present. Economy here implies an objective involving money; money (in
total or as a function of time) may also be present as a constraint.
Objectives and constraints, where mentioned below, relate to the project
unless otherwise noted; the end-product problem is not considered in depth
in this book. Carmichael (1981) considers the end-product problem.
Planning controls are selected over the full time frame of a project for
a globally optimum solution. Planning separately on a stage-by-stage or
even day-by-day basis is suboptimal. Control selections early in a project
constrain future updated control selections.
u(ii is selected based on whatever objectives and constraints are present.
Frequently a dominating constraint is that of a desired project performance
or behaviour (later referred to as state x(ii) over the project duration, i
S
≤i ≤
i
F
; such a constraint may eliminate most of the potential for optimisation.
The selection implies knowledge of work practices, and (fore)thought as to
what might happen later in time. In this last sense, planning is said to be
proactive; the alternative is to be reactive and not carry out forethought
as to what might happen. x(ii has multiple parts (that is, it is a vector
quantity of state variables) that contain information on cumulative resource
usage, cumulative money usage or cumulative production or equivalent.
(Project output, z(ii, is also an indicator of project performance, but for
the stripped-back project formulations given later, output and state are the
same.)
Flow-on aspects to assist management
This book pushes a rational and structured approach to planning as being
preferential to an ad hoc approach.
Planning well-done gives rise to many flow-on aspects that assist the
overall project management. For example:

Critical/important work items can be highlighted. Attention can be
focused on those activities that are critical in terms of achieving a
desired project completion date, or Pareto in some other sense.
Introduction 11

All activities necessary for the project are delineated.

Potential trouble areas and obstacles are identified.

Communication between project personnel can be assisted. Project team
members and other project participants can be informed of their duties
and when they are to contribute to the project and for how long.

Planning outcomes can provide a basis for coordinating the work of
project participants.

Coordination between simultaneous projects can be facilitated.

A more satisfying work environment for project personnel can be
obtained where it can be seen that proper planning has taken place.
Stakeholders can be impressed.

Uncertainties affecting the project can be reduced. Allowances and con-
tingencies can be reduced. Risk management practices are implicitly
incorporated, without needing to mention the term ‘risk management’,
or to separately go through the risk management process. Predicting
the influence of changes and delays is enhanced.
Flow-on aspects to please the project owner
Project owners prefer planning well done because of the improved possibility
of a project finishing by a desired date and within a given budget. Owners
can be provided with information regarding dates and costs (throughout a
project, together with that at project completion), information necessary to
conduct their total affairs. The impact of changes, delays and uncertainties
associated with completion dates can be ascertained.
Case example – Desktop computer software
One organisation, with branches distributed over a large geographic
region and in a number of countries, decided to standardise on an
operating system and applications used by staff on their desktop
computers. A consultant was contracted to provide the redesign and
specification. One particular personal computer brand was selected
as the hardware.
The branch offices were geographically dispersed and the type of
problems differed from country to country, making planning difficult:

Additional funds had to be allocated, taking into account techni-
cal uncertainties and equipment availability; for example a com-
puter may cost twice as much in one country as another.

In some countries, additional funds were required to ensure cer-
tain officials approved of the private network infrastructure.
12 Conventional treatment from a systems perspective

While performing the installation of the equipment, there were
intermittent power failures that resulted in the installation having
to go back to the start of the procedure.

Some local vendors caused some logistics problems that delayed
the project schedule. A vendor would promise equipment to be
delivered based on a contract, but then not do so.

Delays were common in some countries.

Certain specified hardware and software was unavailable in some
countries. Equipment had to be sourced elsewhere. An additional
time period and cost was involved.

Even if the equipment arrived on schedule, unforeseen problems
occurred during the installation of the servers.

No provision was made for additional parts/spares for replace-
ment. This caused delays, which meant that other sites had to be
rescheduled as well.

There was a lack of skilled technical personnel. This required staff
to be retrained in order to do the implementation and provide
support after implementation.

Additional staff had to be sourced from outside the organisation
to assist in the implementation.
In hindsight, an additional time period and cost should have been
included in the planning for this project. With all the problems
encountered, an additional year and a half was required for the project
over that originally planned – getting the infrastructure installed,
equipment delivered and staff trained.
Replanning
Generally the term ‘control’ is used loosely in the literature and by lay per-
sons and practitioners. Particularly at fault is the usage in some computer
packages, where package marketing promotes the feature of ‘control’, but
really refers to something else, in many cases something no more sophisti-
cated than handling data. In other cases the term is used to mean placing
some limit or containment on something such as costs, or keeping this
something from growing unacceptably. In general management texts, and
lay terms, the term ‘control’ is used in the context of power, dominance or
influence. (Something that is ‘out of control’ is unable to be influenced.) It
appears that the word ‘control’, like planning, is used by most people in
the sense of Alice in Wonderland:
‘When I use a word,’ Humpty Dumpty said in a rather scornful tone,
‘it means just what I choose it to mean, neither more nor less’.
Introduction 13
‘The question is’, said Alice, ‘whether you can make words mean so
many different things’.
(Lewis Carroll, ‘Through the Looking Glass’, Ch. 6)
In this book, the term control is used in a systems sense, meaning the
actions, input or decisions of the planner (decision maker) in order to bring
about a desired project (system) performance.
The popular usage of control as stopping something from growing unac-
ceptably or containing something is incorporated through the use of con-
straints in the synthesis formulation.
The common terms ‘project control’, ‘project control system’, and
‘cost control system’ are not favoured in this book because their lay
meanings have a tendency to mislead; their sloppy usage is inconsis-
tent with the optimum control systems usage preferred in this book.
Rather, the term replanning is preferred to ‘project control’.
What replanning, when necessary, does is come up with revised values of
control variables, u(ii, for the remainder of the project, i
now
≤ i ≤ i
F
. The
nature of u(ii is unchanged – multiple parts (vector quantity) that contain
information on method, resources and resource production rates.
Replanning does what is essentially embodied in Bellman’s principle of
optimality, which was devised for different purposes. The principle states:
‘An optimal policy has the property that whatever the initial state and initial
decision, the remaining decisions must constitute an optimal policy with
regard to the state resulting from the first decision.’ The words ‘decision’
and ‘control’ may be interchanged. See Carmichael (1981, 2004).
In essence, replanning is no different to planning. It is planning carried
out subsequent to any original planning, using updated initial conditions,
shortened time horizon, updated constraints and possibly updated objec-
tives. As such, most of the discussion on planning applies to replanning.
Project performance (the project state x(iii, is monitored on an ongoing
(usually regular) basis, compared with planned performance (typically in a
report) and, if necessary, corrective action is taken to send the project in a
desired direction (on a desired-state trajectory). The controls bringing about
corrections may be in the form of method (including sequence), resources
(or money), and/or resource production rates (or equivalent) input to the
project. The controls are chosen to extremise the objectives while satisfying
any constraints present. Applying changed or different controls amounts
to replanning. Accordingly, planning (in the form of replanning) might be
14 Conventional treatment from a systems perspective
referred to as an ongoing or evolutionary process. The ‘planning phase’ of
a project occupies the total project duration.
Generally, it can be said that prior work done constrains future (updated)
planning actions. Hence the constraints are changing as a project progresses.
The flexibility in choice of controls decreases with time.
Monitoring and reporting are necessary precursors to knowing whether
replanning is necessary or not. A changing project and its environment (rep-
resented later by disturbance, i(ii) lead to the need for continual monitoring
and reporting of project progress. Work rarely is done exactly as planned.
(Project) scope also rarely stays constant from project start to project end.
Objectives and constraints may also change.
Planning is based on certain assumptions, and at project implementation,
conditions might be different. As well, it is rare that any project as originally
planned is brought to fruition. For example:

There are inevitable changes relating to technical specifications, method
of implementation and owner needs (revised priorities, rework, . . . );
the scope (of work) changes; variations occur.

The original planning may have been unrealistic, unachievable, inaccu-
rate or contained mistakes particularly with regard to activity duration
estimates, activity relationships and activity resource demands and pro-
duction; activity sequencing used for planning may have been incorrect;
duration and/or cost estimates may have been too optimistic/pessimistic.

There may be vendors/suppliers or contractors who are unable to meet
their original target dates, or perform to expectations/requirements;
subcontractors may be unknown.

Extreme unexpected events may occur including unimagined geotech-
nical conditions; there may be delays due to industrial trouble and bad
weather.

There may be changes in design; the design may not have been checked
thoroughly.

There may be changes in the availability of resources (people and equip-
ment); equipment may turn out to be not suited to the task; sufficient
resource numbers are unavailable when required.

There may be changes in the availability of money.

Material prices may change; there may be changes in the availability of
materials.

There may be technology changes or technical difficulties; the use of new
or previously unused technology may not have been thought through
fully; quality or reliability issues occur.

There may be coordination problems with other groups; necessary pre-
ceding activities may be incomplete.
Introduction 15

There may be access, space or logistics problems.

The market, the economy, competitors, regulations and interest rates
may change.
And according to Murphy’s Law: Anything that can go wrong, will go
wrong.
Case example – Bulk concentrated food mix
The project involved developing a new plant to manufacture a bulk
concentrated food mix for industrial use. The capital component of
the project included civil work comprising of a new building to locate
a bulk sugar storage and handling plant and an addition of a mezza-
nine floor to house machinery in an existing building. The mechanical
work comprised the installation of product silos and associated valv-
ing and pipework, installing new processing and packaging machinery
and installing a new automatic cleaning system. The electrical work
comprised the electrical installation for the above work and imple-
menting software for the automatic control of the mechanical plant
via programmable controllers and computer screens. The total bud-
geted cost of the project was several million dollars but overran by a
third. The time span for the project was estimated to be about a year,
but also overran by a third. The civil work was managed by a pro-
fessional project management company and executed by contractors.
The mechanical and electrical design and installation was managed
by in-house project engineering staff.
This project had both overruns in cost and duration. There were
uncertainties in the planning that resulted in delays in the implemen-
tation stage. These are described below.
In the initial planning, there was no clear estimate of the required
capacity for the new plant. When the project was initially budgeted,
the return on investment was calculated on sales figures based on
potential in the market, that is subjective estimates. The plant was
sized based on the forecast budgeted sales figures plus a growth fac-
tor. However, later in the project, the plant was required to be of a
significantly larger capacity. This affected the plant layout, and the
detailed design could not begin until new dimensions, power require-
ments and prices for machinery were obtained. The delivery periods
were also longer for plant and equipment.
There was a resulting problem in the location of the sugar store and
handling equipment. The architectural design for the building and the
layout had been done based on the original estimates, but there was
16 Conventional treatment from a systems perspective
inadequate space for the new plant size. This resulted in further delays
in redesigning the building and obtaining development approval from
the local authorities.
There was also no certainty in the production process because the
production technique had to be changed for commercial and con-
tractual reasons. The plant was basically of an unproven or novel
design that was scaled up from the research and development (R&D)
laboratory trials. During the commissioning stage, a number of modi-
fications had to be done to achieve the required product specifications.
These modifications were done on a trial-and-error basis and conse-
quently added to the increase in the duration and cost of the project.
There was no project manager for the entire job. The person respon-
sible for the project was the project engineering manager, but he was
supervising a number of other projects at the same time. There was a
project management company that looked after the civil construction,
while three mechanical engineers and one electrical engineer under-
took the design and the supervision of the plant installation and com-
missioning. Again, these engineers were associated with other projects
at the same time. Some sections of the project were completed well
in advance, while others had significant delays mainly due to long
delivery time periods of some plant and modifications to the design.
Case example – Land development project
In a land development project, activities included work in dredging
(tidal river), dewatering bunds, dry bulk earthworks, bridge piling and
piers, rock foreshore protection, sewerage and stormwater drainage.
The original planned construction duration was about a year.
In the initial planning, uncertainties recognised included those asso-
ciated with: weather and flooding, encountering marine clay lay-
ers during dredging and dry excavation, time periods for sequential
approvals by government departments of various components of con-
struction, and contractor performance relevant to schedule and cost.
Undefined material in dredging The chance of encountering marine
clay layers, unable to be dredged, created an uncertainty in the project
duration and cost. In the initial planning, the presence of mud was
not recognised. Available boreholes represented samples some 200 m
apart, and previous experience from construction in this particular
area had indicated that no dry marine mud would be encountered and
Introduction 17
was not allowed for. As the project progressed, a significant quantity
of mud was found in one area. The response to this was to adopt an
alternative method of material removal. Material was trucked, in lieu
of dredging, at an additional cost and an additional duration.
Approval time periods The uncertainty of the approval time period
of subactivities by local authorities was recognised in the initial plan-
ning. However, the duration of the delays in achieving approvals
was grossly underestimated. Changes in environmental legislation
had placed more stringent requirements (for example, additional
groundwater testing) since commencing the project. For this particu-
lar subarea, access was not available until approval had been issued.
This influence was handled by arranging the contractor to carry out
advanced work in other approved areas for future stages. In this man-
ner its resources were not left idle (a potential cost to the project with-
out production). Although achieving meaningful work in the future
stages, the owner’s cash flow troubles caused considerable internal
management difficulties.
Contractor performance The production performance of the con-
tractor was well known for this type of work, and the estimated
activity durations were reflective of expected performance. In the con-
struction phase, there were no difficulties in this respect. While the
tendered project duration appeared reasonable to the consultant and
the owner, there still remained uncertainty in the activity duration
estimates. This was particularly relevant to delays from a flood event.
Information from monitoring, a comparison of actual project performance
with planned project performance, as well as absolute project performance,
is processed into progress reports, which are used as the background to
replanning (Figure 1.5).
One useful report is that which shows together all variances, that is
the difference between as-planned performance and actual performance
for all indicators of project performance (that is, state). A generic form
of collective variance reporting is shown in Figure 1.6. Figure 1.6 shows
a two-dimensional state space (a space with state variables as axes), but
the same holds true for where the performance of the project is reflected
in more than two states; the space becomes a hyperspace for dimensions
greater than three. (For dimensions greater than three, tabular reporting of
these variances, or pairwise reporting of these variances might be preferred.)
Two state trajectories are shown in Figure 1.6 – one corresponding to the
as-planned behaviour, and the other corresponding to the actual behaviour.
The variances in the states are also shown.
18 Conventional treatment from a systems perspective
Implement
Monitor Report
Absolute and comparative
performance
Controls
Plan/Replan
Future
inf luences
estimate Disturbance
Figure 1.5 The replanning process.
As-planned
Actual
State space
Variance in state 1
Variance
in state 2
Trajectories
State 2
State 1
Planned project state
at reporting date
Actual project state
at reporting date
Figure 1.6 Generic way of representing project variances in a state space.
As an example, the two states in Figure 1.6 might be cumulative resource
usage and cumulative production. The variance in cumulative production
will indicate whether a project is ‘ahead/on/behind schedule’; production
performance not only reflects actual resource production rates, but may also
reflect delays, and popular reporting may be more accustomed to schedule
and delay reporting rather than reporting as in Figure 1.6. The variance in
cumulative resource usage, when converted to a money unit, will indicate
whether a project is ‘over/under budget’. Popular reporting may also be
more accustomed to cost reporting rather than reporting as in Figure 1.6; in
Introduction 19
that sense it might be preferred to aggregate the cumulative resource usage
states across all resource types and express this as a cumulative cost state.
The variances from Figure 1.6 might then be summarised over time in
diagrams such as Figure 1.7 and 1.8 to give a historical record of project
performance.
A way of reporting variances, less useful than Figure 1.6, is shown in
Figure 1.9, where individual states are plotted against time. This only gives
a part picture of what has happened on the project up to the reporting date.
For example, where cumulative resource usage is the state being represented
in Figure 1.9, the figure may indicate that less resource numbers have been
used than had been planned, but says nothing as to whether production is
as-planned, ahead of as-planned or behind as-planned.
+
+


Date 1
Date 2
Date 3
Date 4
State 1
variance
State 2
variance
Figure 1.7 Progressive plot of state variances.
Date
Date
+

+

State 1
variance
State 2
variance
Figure 1.8 Separate plots of state variances.
20 Conventional treatment from a systems perspective
As-planned
Actual
Trajectories
State 1
Time
Variance in state 1
Reporting
date
Figure 1.9 A way of showing project variances less useful than Figure 1.6.
Earned value thinking introduces a modification to Figure 1.9 to make it
more useful (when money and production/schedule are being looked at),
but still Figure 1.6 is a preferred form of representation.
Example replanning controls
The replanning problem, in principle, is no different to the planning prob-
lem. The available controls are the same in both cases; replanning may
differ in that it becomes an adjustment or change to previously selected
controls from the earlier planning. Possible replanning controls are many,
ranging from altered manning numbers through bonus systems, preven-
tative maintenance etc., but commonly centre around method, the use of
resources (implying money), and resource production rates (or equivalent),
for example:
Project level controls

Resources may be redeployed, redistributed, increased or decreased
within the project. This could be expected to alter the critical paths and
the scheduling of activities and costing.

Using new/alternative methods (changed work method), and resources
for the remainder of the project. This implies a new network and new
costing.

Fast-tracking the remainder of the project; activities done in parallel
rather than in series; overlapping relationships between activities.

Expediting materials; introducing ‘just-in-time’ or similar inventory
practices.

Outsourcing part of the work.
Introduction 21
Activity level controls

Compression (shortening durations) of activities leading to network
compression

Splitting activities

Trading off resource numbers and duration

Changing the resource work calendars.
Element level controls

Working extra shifts

Using extra resource numbers.
Constituent level controls

Providing incentives

Altering resource production rates.
This list is not totally exhaustive. Controls selected depend on the plan-
ner’s experience and expertise, as well as the situation and some creativity.
Some solutions may be better than others. The best solution is that which
extremises the objectives while satisfying any constraints present. How-
ever, generally planners are after a satisfactory solution rather than some
theoretically optimum solution.
Work study and value management (including constructability/
buildability) studies can be useful here, though all these are subsumed by a
systematic approach to problem solving (Carmichael, 2004).
Replanning may also involve the removal of constraints (physical, tech-
nical, logic, . . . ) affecting the work.
Through all planning and replanning discussions above, it is assumed
that estimates and assumptions are correct, which may not be the case.
Correcting these will also affect the solution.
Planning effort
As with many matters involving effort and associated cost, there is a balance
between overdoing and underdoing. Little planning saves money and effort
on the planning work but may be wasteful in terms of the total project
outcomes (extra cost, late completion, damages, . . . ). Further effort and
money put into planning may improve the project outcomes. Still further
effort and money put into planning may, however, bring no additional
benefits or savings in total project costs.
Unfortunately there is no reliable data on the cost (and amount) of
planning. One reason for this is that people cannot agree on what actu-
ally constitutes planning and where the boundary between planning and
22 Conventional treatment from a systems perspective
other activities should be drawn. Sloppy usage and understanding of the
term ‘planning’ by many people contributes to this. The cost (and amount)
will also vary from industry to industry and organisation to organisation.
Smaller projects may bear a proportionally higher planning cost than larger
projects.
There is even less data on the relationship between planning effort and
project outcomes. Carmichael (2004) argues:
A project contains an immense number of influencing factors ([matters]
that contribute to bringing about any given result), none of which is
possible to be isolated in order to establish its influence on a project’s
outcome. On any given project some factors may be present, some may
not be present, while some factors may be more dominant than others.
It is not possible to demonstrate in any objective way how a partic-
ular factor contributes to project success. (. . . the inverse investigation
problem cannot be solved.) There is only experiential observation or
anecdotes of projects to support popular contentions. To that could
be added ‘gut feel’, judgement and perhaps logic; the contributions
of various factors seem reasonable propositions. Experience says that
the contribution of particular factors reduces/increases the chance of a
project under-performing.
Objectivity is not possible because, although many projects are sim-
ilar, no two projects are the same. A ‘control’ project cannot be set
up, against which other projects with different factors are compared;
the closest that may be obtained is to observe two similar projects but
with different factors, but this is inconclusive. Each project cannot be
done multiple times – each time with a changed factor. Medical exper-
iments on people, it might be argued, suffer the same problem, but
get around the issue to a certain extent by conducting trials on large
numbers of people. Conducting trials on large numbers of projects,
each containing multiple factors that have the ability to influence the
projects’ outcomes, is not feasible. Individual factors cannot be varied
one at a time, while all other factors are kept constant. There is not a
one-to-one relationship between any project inputs/factors and project
outputs; to a certain extent, a project (as a system) is almost uncon-
trollable in systems engineering (Kalman controllability) terms; altering
any one input/factor affects multiple outputs. There are also observ-
ability (in the sense of Kalman) troubles. The benefits purported to be
due to one factor may be due solely or partly to some other project
factors.
Performing one project [task] affects other [tasks], and the outcome
of other [tasks]. And on large projects there are large numbers of [tasks],
most of which are interdependent.
Introduction 23
An alternative is to approach the argument from the double negative
perspective, namely by reasoning that with a certain factor, then project
outcomes will almost certainly be undesirable. And this is supported by
many case projects that have performed badly with such a factor. The
consequent conclusion is drawn. Again this is reasonable, but can’t be
proven.
A number of studies have been reported trying to demonstrate a link
between factors and project success. Typically they involve examining
multiple real projects, and not any deliberate experimental projects.
Projects are categorised, in a pseudo-rigorous fashion, as high, medium
or low in terms of the factors and the (falsely drawn) conclusion then
follows that there is a positive correlation between a factor and ultimate
project success. The studies give an air of objectivity, but remain largely
subjective.
It is difficult to establish what is the optimum (the balance between the
amount of planning and project outcomes) planning effort. It is not even
safe to observe projects and assume that practitioners have homed-in on
the optimum by trial and error.
Case example – Food production
It was common practice within one food production company to
undertake projects with much fervour and zeal. Often projects were
fast-tracked for no apparent reason. Projects, which resulted from
operational or competitive necessity, were also done at such a rapid
rate that all conventional planning and project staging methodologies
were ignored. The bottom line result of this was ‘over budget, over
schedule’ projects, the bane of any project manager.
Engineering-based projects evolved from a production department,
usually the production manager, with inappropriate costing, little
associated planning and a timescale that could not be adhered to.
The shortfall in planning effort also reflected the insufficient knowl-
edge or experience about the job and translated into a poorly con-
structed job specification that led to a shallow scope (of work) and
ultimately unrealistic tender documents.
People within the accounting section at the company often argued
that thorough planning was costly and could not be justified. It was
felt that thorough planning also led to an increase in project duration
and was only a way for some employees to ensure their employment.
This belief, in some ways, promoted the use of fast-tracking in order
to circumvent thorough planning.
24 Conventional treatment from a systems perspective
Project close-outs
Special planning commonly goes into the commissioning and handover part
of the termination phase of a project.
Work in this phase ties together a lot of loose ends, prepares for the
transition to take the facility to the operating and maintenance stages,
analyses project performance, files all project data, disposes of or transfers
resources including people and closes accounts.
Project personnel are aware of the need for thorough planning in the initial
stages of a project but some neglect the planning of the project termination.
Some projects engage a new project manager and team specifically to
cater for the commissioning and handover. This is then viewed as a ‘project’
in itself and planned accordingly.
Personnel problems, as the project winds down and people are transferred
or laid off, are of particular importance during project close-outs.
Book outline
The book develops the subject in two parts:

Part I – Conventional treatment from a systems perspective

Part II – Synthesis treatment.
The reason for this is that the systems ideas presented in the book are new
to planners. Part I takes planning practices (both good and bad) famil-
iar to most planners, but couches them in the more rational framework
offered by systems thinking. This elevates the practice of and thinking on
planning. Part II offers new theoretical constructions for the planning prob-
lem, and although not intended to be used in everyday planning practice,
offers considerable insight into what planning is. Planners will benefit by
understanding true synthetic approaches. Selected topics in both Parts take
a number of interesting planning topics and develop these in a fashion con-
sistent with Parts I and II. Examples are included throughout the book to
reinforce the subject matter.
Figure 1.10 shows a schematic outline of the flow of the book. Part II
could be omitted in a first reading. Part I relies on the reader having some
fundamental exposure to planning.
Terminology uses
Terms used interchangeably by writers include owner, client, principal,
employer, developer, proprietor and purchaser. Generally, the term ‘owner’
is the one adopted in this book.
Introduction 25
1. Introduction
Part I
Conventional treatment
from a
systems perspective
2. Systems thinking
3. Planning tools and terminology A–Z
4. The planning process
5. Planning over levels
6. Replanning
Part II
Synthesis treatment
9. Synthesis over levels
8. Multistage planning
7. Selected topics
10. Selected topic
Figure 1.10 Outline of book.
Terminology used interchangeably in this book includes:

State, output, performance, response and behaviour

Control, action, decision and input.
Gender comment
Commonly references use the masculine ‘he’, ‘him’, ‘his’ or ‘man’ when refer-
ring to project personnel possibly because the majority of project personnel
have historically been male. However, such references should be read as non-
gender specific. Project management is not an exclusive male domain.
Influential publications
The following books by the author have ideas that were influential in the
formation of the planning ideas advanced in this book:
Structural Modelling and Optimization, Ellis Horwood Ltd. (John Wiley
and Sons), Chichester, 306pp., 1981, ISBN 0 85312 283 0.
26 Conventional treatment from a systems perspective
Engineering Queues in Construction and Mining, Ellis Horwood Ltd. (John
Wiley and Sons Ltd), Chichester, 378pp., 1987, ISBN 0 7458 0212 5.
Construction Engineering Networks, Ellis Horwood Ltd. (John Wiley and
Sons Ltd), Chichester, 198pp., 1989, ISBN 0 7458 0706 2.
Contracts and International Project Management, A. A. Balkema, Rotter-
dam, 208pp., 2000, ISBN 90 5809 324 7 / 333 6.
Disputes and International Projects, A. A. Balkema, Rotterdam, 435pp.,
2002, ISBN 90 5809 326 3.
Project Management Framework, A. A. Balkema, Rotterdam, 284pp., 2004,
ISBN 90 5809 325 5.
Exercises
1. What is planning to you?
2. Look around you. The presence of planning is everywhere. Con-
sider how people’s routines are affected by the following end-products of
planning:

What would you do without television, radio and movie pro-
grams/guides?

How would you like to live in a city without transport timetables?

How could you study for a degree without subject, lecture and topic
timetables?

How would athletes achieve peak performance without a training
schedule?
3. Consider a typical project. What form of planning might be under-
taken by the various project participants – owners, consultants, contractors,
subcontractors and suppliers?
4. Lecture timetables, transport timetables, television programs, training
schedules, bar charts, time-scaled networks, and cumulative production
plots are all examples of ways planning information is conveyed. Why do
we use one way in one situation and another way in a different situation?
Examine some timetables, programs etc. and try to improve on the way
the planning information is conveyed.
5. The question many people ask is ‘why plan?’ if plans are continu-
ously changing. Also some people state that they have been associated with
projects where no planning was carried out and the projects were successful.
How do you respond to such comments?
6. List benefits of thorough planning. What are the disadvantages or draw-
backs to thorough planning?
Introduction 27
7. In what way can a plan limit flexibility in developing new ideas, stifle
initiative, and focus people’s attention on the small picture at the expense
of the bigger picture?
8. Rank, in your opinion, the following planning-related reasons why
some projects perform poorly. Use a scale from 1 to 10, with the strongest
cause as 1 and the weakest cause as 10.
– Inadequate period of time and effort spent on planning
– Lack of ability on the part of the planner
– Lack of management support for planning
– Lack of end-user

support for the plan
– Lack of involvement of end-users

in the planning
– Lack of justification of the plan
– Inadequate or non-existent monitoring
– Disregarding the plan once developed
– Monitoring of project performance but no feedback for corrective
action
– End-users

not being properly informed of their duties under the plan.
[

Planning end-users refer to those who are to use the plan, the so-called
‘doers’.]
List planning-related reasons, other than given above, as to why some
projects perform poorly, and rank these reasons as well.
What are the causes of these practices? The use of a cause-effect, Ishikawa,
fishbone diagram or similar might be useful here.
What suggestions can you make in order to eliminate the causes of these
practices?
9. Estimate the cost involved in the planning of a project that you are
familiar with. Base your estimate on the number of planner-hours that you
think are involved.
Estimate the cost of delay (avoidable) in the project’s completion date by
one day, one week and one month. Base your estimate on component costs
such as holding charges (interest on borrowed money), liquidated damages,
overheads, equipment and facility hire, and wages.
Compare the two amounts estimated.
Can the degree of planning observed be justified on an economic basis?
In general, should planning effort be commensurate with the results
required?
10. List possible sources of control that could be used in the follow-
ing situations in order that the project does not diverge further from the
planned state. No change in scope is allowed. (In particular, the specifica-
tion cannot be altered.) No unethical practices are allowed. It is assumed
28 Conventional treatment from a systems perspective
that the original planning and design for the project have been undertaken
in an optimum fashion, and there is no ‘fat’ available for trimming.

The project is running behind production/schedule.

The project is going over budget.
11. Project duration and cost are related. But how are they related? There
is a view put forward on large projects that ‘Extra duration costs extra
money. Reduced duration saves money.’ What is your view?
12. The sale of something can be treated like a project. Imagine you are
selling your car or house. What kind of planning would you undertake?
13. Plot, on the following diagram, your view of planning-related tasks
over the duration of projects, from the point of view of:

The owner.

The main contractor.
Concept Implementation
Amount of
planning-
related
activities
Project
start
Project
end
Termination Development
14. Both planning and design are inverse systems problems. Broadly, in
both problems you are trying to determine the input of a system in order to
achieve a desired output. This is in contrast to analysis where you determine
the output of a system resulting from a given input.
List the conceptual similarities between planning and design.
15. Project close-outs commonly have a heavy emphasis on a fixed com-
pletion date. Planning may be done backwards fromthis point. What special
planning problems would this create compared to planning being done in
a forward sense?
Chapter 2
Systems thinking
Introduction
Framework
Much of this chapter is based on Carmichael (2004). The conceptual frame-
work offered by control systems theory is regarded as the most suitable for
developing planning thinking.
Thinking on planning is developed in two ways (Figure 2.1):

The modelling of projects as control systems, both single level and mul-
tilevel, with attendant project descriptors of control, state and output
variables.

The interpretation of planning as an inverse systems problem, and in
particular a synthesis problem. With project changes and with project
evolution, replanning becomes an evolving set of similar inverse prob-
lems. With the evolution of a project, the inverse problem becomes
better and better delineated. The same type of problem is being solved
repeatedly throughout the project. Inverse problems by their nature
have non-unique solutions; uniqueness is established within an optimal
control systems formulation.
Projects in the following have been stripped-back or simplified to their
mechanical backbone (Figure 2.2). Issues relating to people behaviour,
which would ordinarily be superimposed on this backbone, are considered
separately.
Systems–subsystems
If a project is regarded as a system, then just as a systemcan be broken down
into subsystems, so a project can be broken down into subprojects. The
decomposition may be continued to sub-subsystem (sub-subproject) level
and beyond. (The process can also go in the other direction, from projects to
30 Conventional treatment from a systems perspective
Modelling
projects as
control systems
Multilevel
Single level
Planning as
an inverse
systems problem
(Optimal)
synthesis
problem
Iterative
analysis
problem
Figure 2.1 Idea development within this book.
Resources
(people,
equipment)
Money unit
Stripped-back
project
People
behavioural
skills and
issues
Figure 2.2 Shaded portion showing stripped-back project issues considered in this
book.
‘program’, a ‘program’ being regarded as a collection of projects.) Decom-
position can be carried out in a number of ways, for example Figure 2.3. The
choice of decomposition will suit the intended project usage. In Figure 2.3,
for ‘system’ read ‘project’, in terms of later chapter developments.
With the example decomposition of Figure 2.3a, the subsystems are all
characterised similarly. With hierarchical decomposition of Figure 2.3b,
successive decomposition leads to finer detail, and a change in the informa-
tion type; the terms ‘higher’ and ‘lower’ are used when referring to levels
in such a decomposition. (Figure 2.3a represents a two-level hierarchy even
though it is called a staged (phased) decomposition.)
Commonly, project decomposition may be according to:

Phase (chronological); time period phasing

Work type or nature. Resource groupings

Work parcels
Systems thinking 31

Contracts

Region, location or geographical area

Organisational breakdown

(Existing) cost codes, cost centres, cost headings

The nature of the planning network. Activity groupings

Function
or generally according to:

Project

Subproject

Sub-subproject, work package, job, . . .

Activity

Element

Constituent.
Subsystem
1
Subsystem
2
Subsystem
N
Interaction
Interaction
System
. . .
(a)
(b)
System
Subsystems
Sub-subsystems
Interaction among
subsystems
Interaction among
sub-subsystems
Lower level
interaction
Figure 2.3 Example system decompositions (after Carmichael, 2004). (a) Staged
(phased) decomposition; (b) Hierarchical decomposition.
32 Conventional treatment from a systems perspective
Subsystems combine through their interaction to give the system at the next
higher level.
Subsystems, sub-subsystems, . . . share the same properties and descriptors
as systems, and hence are ‘systems’ in their own right.
Single level system
To understand systems ideas, it is convenient to start with a single level
system. Multilevel, hierarchical systems (including staging) are collections
of single level interacting subsystems.
Environment
The environment is defined as everything except the system (Figure 2.4).
The environment affects the system by changes, and is affected by sys-
tem changes. Constraints, project initial and final (terminal) conditions,
and disturbance reflect the environment and the environment–system
interaction.
Input and output
A system may be regarded as an input–output pair or input–output transfor-
mation (Figure 2.5). Input and output contain multiple parts (that is, they
are vector quantities of input variables and output variables respectively),
and the singular form of the terms is used to refer to both a single entity
or plural entities, though sometimes the plural terms ‘inputs’ and ‘outputs’
may be used when the plural is specifically intended.
Subsystems and lower level components also have inputs and outputs.
Environment
System
Figure 2.4 System, environment distinction.
System
Input Output
Figure 2.5 System as an input–output transformation.
Systems thinking 33
Input, control
Input is equivalent to a decision, action or control. The term ‘control’ is
favoured here in place of input, decision and action. The planner (decision
maker) selects the project control in order to get a desired project output
(performance, behaviour or response).
The preferred control is that which extremises the project objectives.
Control contains multiple parts (that is, it is a vector quantity of control
variables), and the singular form of the term is used to refer to both a single
entity or plural entities, though sometimes the plural term ‘controls’ may
be used when the plural is specifically intended.
Output, state
The output describes the external or observable system performance,
behaviour or response. (The terms ‘performance’, ‘behaviour’ and ‘response’
are used interchangeably in this book.) Internal system behaviour is usually
given in terms of system state. For the stripped-back projects considered in
this book, output and state are the same, that is there are no observability
(in the sense of Kalman) issues.
Output and state are controlled variables. In planning, it is the project
performance or behaviour which is being influenced by the choice of the
control.
‘State’ contains multiple parts (that is, it is a vector quantity of state
variables), and the singular form of the term is used to refer to both a single
entity or plural entities, though sometimes the plural term ‘states’ may be
used when the plural is specifically intended.
Disturbance
In most systems studies, there is something that prevents the hoped-for
output being obtained exactly. This corruption is represented by disturbance
in Figure 2.6.
Disturbance reflects influences (on project performance) that are different
to that anticipated when (re)planning (and which prevent as-planned per-
formance being attained). This includes the impact of unforeseen influences
System
Input Output
Disturbance
Figure 2.6 System with disturbance.
34 Conventional treatment from a systems perspective
and influences that do not remain constant throughout a project. It implies
uncertainty.
Automatic control – Technical systems
The automatic control of technical (mechanical, electrical, chemical, . . . )
systems may be done in a number of ways. Of relevance to planning and
replanning are:

Open loop control. Information flow is one way. The control is
selected up front, and no follow-up monitoring of system performance
and corrective control or action is carried out (Figure 2.7a).

Closed loop (feedback) control. There is a circular transmission of
information (Figure 2.7b). Closed loop control is commonly used in an
error control form. The control is selected and feedback occurs based on
the difference between desired system performance and actual system
performance.
Closed loop (feedback) control attempts to nullify the effects of disturbance
on system performance, and to keep the system performance as desired.
It has the better chance of ensuring that the system’s performance goes
close to that desired, compared with open loop control where the project
performance may or may not go close to that desired.
System
Input Output
Feedback
Disturbance
Comparison
System
Input Output
Disturbance
Monitor
(a)
(b)
Figure 2.7 Comparison of open (a) and closed loop (b) representation.
Systems thinking 35
In error control, feedback may be in a positive or negative sense. Neg-
ative feedback decreases the deviation between desired and actual perfor-
mance, with the situation of zero deviation being the ultimate target. (Pos-
itive feedback increases the deviation.) There is no feedback if there is no
deviation.
Both open loop and closed loop control include anticipation of the effect
of future events/influences (much like the approach adopted in risk manage-
ment). The control is selected with the view that future system performance
is going to be affected by future events/influences.
The Deming PDCA (plan, do, check, act) cycle is no more than a closed
loop control system.
Planning
Planning uses open loop control incorporating estimates of future influ-
ences. This gives a project ‘baseline’ or ‘target’ performance.
Replanning
Monitoring and reporting of project performance gives project status in
terms of absolute performance and performance relative to the baseline
(errors or variances).
Replanning uses open loop control incorporating updated estimates of
future influences, and initial conditions equal to the current project state.
Error control ideas are used to guide the choice of which controls need to
be adjusted, but are not used in an automatic closed loop control sense in
current replanning practice. Replanning gives an updated project baseline
or target performance. Monitoring and reporting follow, and the cycle
continues throughout the project.
Replanning modifications of control systems ideas include:

Controls are not automatically selected, but rather the result of plan-
ner intervention. The planner adjusts the input (control) to a project
(which may be subjected to future influences) to give a desired project
performance.

Many future events are not foreseeable or foreseen. This introduces
(unwanted) deviations from desired performance, and this in turn can
force a change in scope, work method, resource usage and resource pro-
duction rates. Hence, for error control purposes, there is the possibility
of the planned baseline of desired performance (from which errors are
measured) changing as the project progresses.

The measurement of project performance is done in most situations
with not much accuracy, making precise control selections not possi-
ble. As well, the relationship between control and performance can be
36 Conventional treatment from a systems perspective
complicated because of the innumerable things happening on a project
at any one time.

Without the instantaneous use of information coming from the moni-
toring of the project performance, time lags within the process interfere
with the choice of effective controls. Collected project data may be over-
taken by new events. Control is applied based on an earlier situation
rather than the current situation.

Both open loop and closed loop control are not used in a real-time
sense.

No state estimation is undertaken. For the stripped-back projects con-
sidered here, it is assumed that there is no noise in the observation
device (though this is not true where people are involved in project
monitoring/data collection). In the sense of Kalman, there is no observ-
ability issue. The output equals the state. There is also no controllability
issue.
Fundamental systems problems
In terms of Figure 2.5, the fundamental systems problems are:

Analysis – given the control/input and a model of the system, evaluate
the output/response.

Synthesis – given a model of the system and the output/response, eval-
uate the control/input.

Investigation – given the control/input and the output/response, evalu-
ate the model of the system.
In each, something different is known and it remains to evaluate the
unknown.
Synthesis and investigation are known as inverse problems. In general
the solution of such problems is non-unique, whereas the solution of the
analysis problem is generally unique. Non-uniqueness in the synthesis and
investigation problems might be removed through additionally including
some objectives or equivalent. (Note that the presence of more thanone objec-
tive introduces subjectivity, which is not present in the single objective case.)
Planning and design (and management) are seen to fit within the synthesis
categorisation.
Synthesis requires the a priori specification of a systems model, that is
the relationship between input and output.
Synthesis problem components
The synthesis problem is essentially a converse to the analysis problem.
Given a certain (desired) response and model, the issue is to determine
Systems thinking 37
Model
Objective(s)
Constraints
Optimal
synthesis
problem
Figure 2.8 Synthesis problem components.
the control that produces this response. Generally, a certain response is
realisable with many control choices. That is, the solution is non-unique.
Further requirements such as extremisation (optimisation, maximisation or
minimisation) of objectives, for example minimum duration or minimum
cost, are needed to make the solution unique.
See Carmichael (2004) for a rigorous meaning of the term ‘objective’.
Generally one desires to synthesise a system which is optimal in a certain
sense; objectives (or equivalent terms) resulting from an imposed value
statement are implied.
The (optimal) synthesis problem is seen to have three main components
(Figure 2.8) (Carmichael, 1981):
1. Model
2. Objective(s)
3. Constraints.
These problem components are expressed in terms of control, state and
output variables.
Optimal synthesis
The optimum problem is within the realm of the well-delineated body of
theory and techniques of optimal control systems or optimisation theory.
In this sense the planning problem is a single level or multilevel, single or
multiple objective optimal control problem.
Optimal control theory is found extremely useful in elucidating the plan-
ning problem. Using state and control variable ideas, the planning problem
can be readily exposed and understood.
Insimple terms, synthesis is equivalent tochoosingthe project control; opti-
mal synthesis or optimal control selects the control so as to extremise some
objectives. In addition, supplementary constraints are also usually present.
Optimal control theory has applicability across all technology and
management.
38 Conventional treatment from a systems perspective
Conversion of synthesis to an iterative analysis problem
The solution of the synthesis problem could be expected to be more com-
plicated and more difficult than the solution to the analysis problem. In
some cases the analysis problem can even be solved intuitively, whereas the
solution of the synthesis problem involves more rigorous methodology.
Because of the degree of difficulty of solving synthesis problems compared
to analysis problems, many synthesis problems are solved in an iterative
analysis fashion.
Established planning procedures are iterative in nature. The iterations
arise from the analysis-based mode of attack on the planning problem and
are not inherent in planning. That is, established planning practices use
iterative analysis.
The writer’s preferred analysis mode of attack on problems is through the
body of knowledge on systematic problem solving, or equivalently systems
engineering.
Systematic problem solving goes through the following steps (together
with iterative or feedback loops):

Defining the problem (including issues and situation)

Selecting objectives

Generating ideas, alternatives

Analysing ideas, alternatives

Selecting the best alternatives (evaluation)

Action.
Alternatively, a systems engineering problem goes through the following
steps (together with iterative or feedback loops):

Defining the problem (including issues and situation)

Selecting objectives

Synthesising systems

Analysing systems

Selecting the best alternatives (evaluation)

Action.
See Carmichael (2004).
Feedback occurs between steps in trying to refine the problem(Figure 2.9).
Asystematic approach to the above steps generates clear thinking at each step
and lends objectivity to a procedure which would otherwise be considered
intuitive.
The objectives are evaluated for each set of values of the control variables.
Similarly the performance, behaviour or response is evaluated for each set
of these values. It is hoped that by adjusting the initial guesses for the con-
trol variables, the response resulting from the analysis will become more
Systems thinking 39
Analysis
Is behaviour
satisfactory
or as required?
No
Yes
Are the values of
objectives
satisfactory?
No
Yes
Synthesis
complete
Problem definition
Alternatives – guess
controls
Evaluation
Adjust controls
Objectives, constraints
Adjust definition;
objectives, constraints
Figure 2.9 Iterative analysis version of synthesis.
favourable and also the values of the objectives will improve (decrease or
increase as appropriate). However there is no guarantee of this, although
the experience and knowledge of the person attacking the synthesis problem
will usually head the iterations in the preferred direction. Where this person
has synthesised such systems before, the person’s first guess of values for
control variables maybe close to‘optimum’ andnoiterations maybe required;
nevertheless the synthesis is still being approached via iterative analysis.
Established, experienced planners would be expected to make good first
guesses and thereby eliminate the iterations. Beginner planners, on the other
hand, would be expected to do several iterations.
The end point of the iterative modification with feedback loops is based
upon obtaining satisfactory project performance and satisfactory values for
the project objectives.
Chapter 3
Planning tools and
terminology A–Z
Activity
Also called task. Technical disciplines tend to use the term ‘activity’, while
non-technical disciplines seem to prefer the term ‘task’. The term ‘activity’
is used in this book.
The term ‘activity’ may refer to a physical activity, an administration or
a management activity, a material delivery activity, or anything that has a
finite duration and uses resources (and consequently money); an identifiable
piece of work.
Exceptions to this include:

A dummy activity which exists to maintain work method logic in a
network, has zero duration and uses no resources. A dummy activity
might be referred to as a logic restraint. Project start and project end
activities may also be optionally introduced in order to give identifiable
start and end points to a project.

An artificial activity which has a finite duration but uses no resources.
An artificial activity may be used, for example, where a schedule delay
or constraint is required. Occasionally, certain activities have to be
completed by a prescribed date or within a prescribed time period (a
duration limitation). This constraint may be straightforwardly handled
by introducing artificial activities into a network, where the duration
of the artificial activity equals that prescribed time period.

An aggregated activity, which is a collection or group of activities,
used to give economy of presentation where detail is not required
by the viewer. The terms ‘summary activity’ or ‘hammock’ may also
be used.
An activity occurs at the lowest level of a work breakdown structure; the
level at which detailed duration and resource estimates (and consequently
money estimates) are commonly applied.
Planning tools and terminology A–Z 41
Activities may be summarised for end-user purpose according to:

Description

Associated work method; incorporated materials (including
potential suppliers, origin and freight)

Logic sequence (relationship to other activities)

Specification and contractual related matters

Resource needs
– People (engineering and other services; personnel and organi-
sational responsibilities)
– Equipment (including facilities, manufacturing and origin)

Activity production rate

Duration

Money (cost and income) estimate. Cost account number

Assumptions.
A continuous activity is one that once started continues until complete. An
intermittent activity is one that can be stopped and started with breaks in
between.
Bar chart
Also called Gantt chart and Gantt bar chart (after Henry L. Gantt’s work
in the early 1900s). Technical disciplines tend to use the term ‘bar chart’,
while non-technical disciplines seem to prefer the term ‘Gantt chart’. The
term ‘bar chart’ is used in this book.
A bar chart may be referred to as a program or a schedule. But there
are other diagrams and charts which are also referred to as programs and
schedules.
A bar chart contains information at both the activity and the project
levels. A bar chart has a timescale horizontally (but no scale vertically),
with the activities listed vertically. The horizontal scale is ideally matched
to a calendar allowing for stipulated work hours per day, weekends, public
holidays, rostered days off (RDO), number of shifts per day and so on. The
activity listing may be in any order, though usual preference (to facilitate
project monitoring) and common practice is to list activities in order of their
earliest start times, leading to a stepped appearance from the top left corner
to the bottom right corner. Alternatively, like activities may be grouped. A
horizontal bar is drawn adjacent to each activity’s name showing that activ-
ity’s scheduled duration, between the start and end times of that activity,
42 Conventional treatment from a systems perspective
and float may also be added. Bar charts may be subdivided (horizontal
partitions) into subprojects. Activities might be collected together (as also
in a subproject) and represented by a single aggregated activity.
Activities are preferably shown connected, as in Figure 3.1, indicating
the order or logic of work or activity dependence. Such a diagram may be
referred to as a connected (or linked) bar chart. In such a form, a bar chart
is an activity-on-link diagram drawn to a timescale. It assists both planning
and replanning; the influence on following activities of an activity change
can be readily established.
In Figure 3.1, the scheduled duration of an activity is shown by a hollow
box; float is indicated by a shaded box. Activities are preferably shown
with float.
Although a bar chart can be drawn directly by scribbling horizontal
lines on a page, a more rational way of obtaining a bar chart is via a
network (critical path) analysis because this includes the work method
and float information. The practice of drawing lines on a page (without
going to the trouble of a critical path analysis) to give the impression of
a rationally developed bar chart has little merit. The information from a
network analysis is readily transferable into bar chart form. Activity start
10 20 30 40 50 60 80 Days
Link
0
Sales
Float
Scheduled duration
Detailed design, options
Approval for options
Preliminary design
Approval for base components
Detailed design, base components
Manufacture base components
Assemble base components
Assemble options
Apply finishings
Marketing
Deliver base components
Figure 3.1 Connected bar chart for an example project. See the associated networks,
Figure 3.25(a,b), and time-scaled network, Figure 3.55. (Activity names have
been abbreviated because of space limitations – work items are implied.)
Planning tools and terminology A–Z 43
and finish times and floats are calculated in the network analysis. The
connections between activities reflect the logic of the network diagram.
(Some computer packages might appear to go in the reverse direction [that
is, bar chart to network] but this is not the case – the so-called bar chart
developed first is in fact a time-scaled [activity-on-link] network.)
Figure 3.1 shows the bar chart plotted for all activities starting at their
earliest start times. However, the start times of the non-critical activities
can be varied because of their float. Hence, in principle, a large number
of bar charts are possible for any given project. This flexibility is used by
planners to optimise the usage of resources.
In summary, a bar chart gives information on when activities begin, when
they end, how long they take, the amount of float, and activity dependence
with other activities.
With connections between activities, the bar chart contains the same
information as a network, but usually is more readable. As noted, a con-
nected bar chart is a time-scaled activity-on-link network. When ‘inte-
grated’, a bar chart becomes the same as a time-based cumulative production
plot. Cumulative production plots, bar charts and multiple activity charts
contain essentially the same information.
Bar charts also find use in reporting for replanning purposes, where actual
project performance is included additional to planned performance.
Baseline
A baseline refers to the as-planned situation. It may be referred to as a
target.
The baseline may be with respect to resource usage (as-planned resource
plot, as-planned cumulative resource plot), money usage (budget,
as-planned cash flow), schedule (as-planned program), production (as-
planned cumulative production plot), and delays (as-planned cumulative
delay plot).
Actual or as-executed progress and performance is compared with the
various baselines for resource types, cost, production and so on, as part of
reporting.
Budget, budgeting
Refer ‘financial planning’.
Cash flow
The difference between income (inflow) (payments received) and expendi-
ture (outflow) (payments to suppliers, employees, trade contractors, interest
on own or borrowed funds, . . . ) is the cash flow, which can be shown in
44 Conventional treatment from a systems perspective
a cash flow diagram or a table. A positive cash flow indicates that more
money has been received than has been paid out; a negative cash flow
indicates the opposite situation.
Closed loop control
Closed loop control may be referred to as feedback control. It is commonly
used in an error control form; the control is selected based on the difference
between desired system behaviour and actual system behaviour.
Compression
Project compression refers to the reduction or shortening of a project’s
duration. The project compression problem is a subproblem of the total
planning problem, where emphasis is placed on the different level controls
that can lead to activity and project duration shortening, and all other
controls are held fixed.
As part of an iterative analysis attack on planning, project compression
may be considered:

Where the calculated project duration, based on normal (least cost)
durations, is excessive.

In order that the overall project completion time meets some required
completion time or deadline, perhaps where penalties or damages for
completion time overrun start to accrue. A similar situation arises
should there be bonuses for early completion.

When, part way througha project, the project is running behindschedule.
Material, equipment or worker/personnel shortages may develop dur-
ing a project thereby delaying the project completion time. Delays due to
other causes (suchas those due toweather, industrial disputes, unforeseen
workplace conditions, . . . ) may occur during project execution.

To assist resource and money allocation on a project or over several
projects, it may be desirable to complete a project early in order to,
say, reuse the equipment or workers/personnel on another project.

Where there may be large time varying indirect costs, decreasing the
project duration may give a lower total (direct plus indirect) cost solu-
tion. See ‘estimates’ for a distinction between direct and indirect costs.
Aproject’s duration may be reduced by reducing or shortening the durations
of critical activities (activity compression).
Compression may come at a cost penalty.
Reducing the durations of non-critical activities does not reduce a
project’s duration, but introduces more float, and might be considered as a
possible way of utilising resources more effectively.
Planning tools and terminology A–Z 45
In the compression of activities, attention focuses on what are termed
normal durations and costs, and crash durations and costs. Activity normal
durations are those of least direct cost and are usually the durations for
which the initial planning calculations are done. Durations cannot go below
activity crash durations – they are the absolute minimum activity durations
obtained by employing maximum resource numbers (Figure 3.2). Crashing
consequently refers to taking compression to the limit. A theoretical con-
struct called a cost slope can be used to rank activities and establish an
order for compressing. A cost slope may be defined as the relative increase
in cost per unit time saved undertaking an activity. Activities with the lower
cost slopes are chosen first for compressing because these lead to lower
increases in total cost. The order of compression is selective. The use of
such an approach in practice will depend on having adequate estimating
data, and this may not be available.
Figure 3.2 represents the conceptual form of activity cost–duration data.
Generally, however, the continuous range of cost–duration information
shown in the curve is not available to planners. Instead, estimating, manu-
facturers’ or contractors’ information may be limited to only two or three
cost–duration data points as shown in Figure 3.3. The number of possible
ways of undertaking an activity may also reduce the data to a few (finite
number of) points. Cost–duration data points represent the minimum costs
to perform the activity in the given durations. In such cases, where it is only
possible to undertake an activity in a finite number of ways, for example by
using a team of three workers or a team of five workers, cost–duration data
points referring to situations in between these defined points are meaning-
less. As another example, consider a delivery activity; the item or material
might be delivered by air within a short time period, but high cost, or by
Crash
cost
Normal
cost
Activity duration
D
i
r
e
c
t

c
o
s
t
Uneconomical
drag out
Further crashing
of activity
Crash
duration
Normal
duration
Cost slope
=
increase
in cost/decrease in duration
Figure 3.2 Compression terms.
46 Conventional treatment from a systems perspective
Crash
cost
Normal
cost
Activity duration
o
o
o
Changing
cost slopes
o Available data
Crash
duration
Normal
duration
D
i
r
e
c
t

c
o
s
t
Figure 3.3 Available activity cost–duration data.
sea at less cost but longer time period. These two modes of delivery define
two cost–duration points. The interval between these two points is not
defined. Any assumption relating to a continuous cost–duration trade-off
is not relevant.
However, the treatment of this discrete case is similar to that in which a
continuum of possible cost–duration data points exists (continuous case).
A special form of the discrete case is when the normal and crash points are
the same; that is the cost–duration data is confined to one point and there
is only one way of undertaking that activity.
A variant of the cost–duration diagram of Figure 3.3 occurs when there
are two or more distinct ways of carrying out an activity. Figure 3.4 indicates
the idea where there are two methods available, and hence two distinct cost
slopes depending on the method used.
A number of writers raise the issue of the validity of assumptions relating
to continuity and linearity of cost–duration curves. A possibly more serious
issue is the (un)availability of multiple cost–duration data for any activity,
and the accuracy of this data.
Compression without activity cost data can also be carried out, but here
the choice of which critical activities to compress becomes almost arbitrary.
There are many ways that activities can be selected.
Project compression may be viable because, even though the direct costs
of activities increase and hence the direct cost of the project increases as the
project duration is shortened, time varying indirect costs are less for shorter
duration projects. There is consequently a trade-off between increased direct
cost and reduced indirect cost (Figure 3.5). If there are large time varying
indirect costs, decreasing the project duration may give a lower total (direct
plus indirect) cost solution.
Planning tools and terminology A–Z 47
Activity duration
Method A
Method B
D
i
r
e
c
t

c
o
s
t
Figure 3.4 Example cost–duration data for two work methods.
Time varying
indirect costs
Direct costs
Original completion
date
P
r
o
j
e
c
t

c
o
s
t
Time
Direct costs + Indirect costs
'Optimum cost'
Figure 3.5 Schematic of direct cost and indirect cost trade-off involved in project
compression.
For control selection, in the compression problem, general ways of short-
ening a project that fit within the above discussion include:
Constituent level

Change the resource production (constitutive) relationship

Use a carrot (incentive, reward)/stick approach.
Element level

Use additional resource numbers (quantity)

Work overtime or multiple shifts.
48 Conventional treatment from a systems perspective
Activity level

Trade-off additional resource numbers (quantity) versus activity
duration.
Project level (changing method)

Run concurrent/parallel activities

Use overlapping relationships.
The solution steps to the project compression subproblem are no different
to any project planning problem.
Compression algorithm
Adopting an iterative analysis approach to planning, it is possible by examin-
ing a networkandthe associatedcost–durationrelationships for the activities,
tomanually compress a network, or towrite a computer programthat mimics
the manual procedures. Coarse steps in the process followFigure 3.6.
The compression calculations may pass through several phases. The three
most notable are outlined below:

Compressing of critical activities may proceed to their respective crash
durations and no lower. If the status of other activities has not been
affected during this process, then the compression calculations are com-
plete. (Similar comments apply to the simultaneous compressing of
multiple critical paths.)

Compressing of critical activities may proceed until a non-critical activ-
ity has its float reduced to zero and hence itself becomes critical (and
hence itself a candidate for compressing).

Where multiple (parallel) critical paths exist in a network, then any
compressing of activities has to be carried out simultaneously and by
the same time interval on all critical paths. Compressing on less than the
full number of critical paths will not reduce the total project duration.
When no further compression is possible, the process is complete.
Some oddities that may occur in connection with network compression
are as follows. Often compressing of one activity which leads to overtime
being worked on that activity may lead to other activities also being carried
out on overtime purely to maintain industrial harmony in the workplace
even though the project may not require these other activities to be carried
out in other than normal durations. There is also the possibility that in
order to compress one activity it may be necessary to introduce a new piece
of equipment to the project. This piece of equipment may then be used on
other activities to gain maximum use out of it although, again, these other
activities may not require speeding up.
Planning tools and terminology A–Z 49
Starting solution
or all normal solution
(Re)analysis of network
Identify critical
activities
Identify non-critical
activities
Rank in order of increasing
cost slope, taking into
account parallel paths
Select activity(ies) with
lowest cost increase and
which can still be compressed
Compress by 1 time unit
Figure 3.6 Steps in project compression.
Computer packages
There are many computer packages of varying user-friendliness available
on the market to assist project planning. They perform the analysis part
of the iterative analysis mode of attack on planning. They may advertise
‘control’ capabilities, but this is ‘control’ in a lay person’s loose meaning
of the word. They treat planning subproblems such as resource smoothing,
resource constrained scheduling and project compression, but in replanning,
at best they only provide information on which replanning is based.
Computer packages based on the critical path method of analysing net-
works have been around for several decades now. The early packages
operating on mainframe computers required punched card input and batch
processing with slow turnaround time period between computer runs. This
was replaced by mini computers and paper tape input. The latest com-
puter packages exploit the graphics capabilities of personal computers, and
involve editing directly on the monitor and almost instantaneous results
for the user. (However, many established planners still prefer to draft the
project network by hand on paper before transferring it to the monitor.)
The latest packages are also considerably more sophisticated in the num-
ber of options they allow and their computer input and computer output
capabilities.
50 Conventional treatment from a systems perspective
Several decades ago, it was not uncommon for an organisation to have a
computer program especially written for scheduling and reporting. All pro-
grams were based on the critical path method. Journals of that era published
listings of simple computer programs. There arose a number of commercial
packages from this background, generalisations of organisations’ in-house
packages, in an attempt to tap a larger market and get a return on the
invested resources writing the computer programs. Some packages became
well known, but all of that era seem to have disappeared and have been
replaced or upgraded with more user-friendly packages, in line with the
more user-friendly graphical interfaces of modern computers.
Today, there are many packages in the marketplace, varying in their price
and user-options. Both activity-on-node and activity-on-link formats are
catered for. Organisations also supplement these planning packages with
word processing, spreadsheet and database packages. Few organisations
write their own software any more. Most use off-the-shelf packages and
adjust their internal procedures to the limitations this imposes.
Frequently, the choice of package that anorganisationuses is dictatedbythe
package requestedby the client, andsosome adaptability of thinking between
packages is required. Fortunately, the packages tend to have common bases.
Planners need to be careful not to fall into the trap of accepting without
questioning anything that comes out of a computer analysis. There is the
acronym GIGO (garbage in, garbage out) and of course Murphy’s Law
(extended) has something to say on the matter: To err is human, but to
really foul things up requires a computer.
A suggested approach is to always interact with any computer package,
leaving the routine analysis calculations to the computer package but not
the decision making (control selection).
Constituent
The basic level in a project hierarchy, referring to resource behaviour.
Constraints
The plural terms ‘objectives’ and ‘constraints’ are generally used in this
book even though the singular forms may be applicable in some situations;
the exception is where the singular is deliberately intended.
All projects have genuine constraints such as funding, environmental
(natural) and political constraints. Constraints restrict the range of control
choices possible.
As with objectives, many people confuse a project constraint with an
end-product constraint. End-product constraints may influence project con-
straints.
Constraints may be stated at all project levels.
Refer Carmichael (2004).
Planning tools and terminology A–Z 51
Contingency
Refer ‘reserve’.
Control
Planning establishes the value of control throughout the project duration.
Control has multiple parts (that is, it is a vector quantity of control variables)
that contain information on method (including sequence), resources (and
hence money) and resource production rates (or equivalent).
The term‘control’ is favoured in this book over input, decision and action.
The planner selects the project control in order to get a desired project
output (response, performance, behaviour). The preferred control is that
which extremises the project objectives.
Critical activity, critical path
A critical activity is usually defined as one with no float. A non-critical
activity is one with float. The critical path is the set of network-connected
critical activities from project start to project end determining the project
duration; it is the longest duration path through a network. If the start or
finish of a critical activity is delayed, the project duration will be extended.
It is possible to have multiple critical paths in a network.
Acritical activity is akintoa Paretoitem, amongst the collectionof all activ-
ities, thoughan80:20distributionof non-critical activities tocritical activities
rarely occurs. It is recommended that most attention be paid to the critical
activities during project implementation in order that the project does not run
over schedule.
The terms ‘critical’ and ‘critical path’ are popularly used terms in the
project management field, often used incorrectly. Network analysis gives
definiteness to these ideas.
Critical path method, CPM
The critical path method (CPM) is based on a deterministic network anal-
ysis, and broadly is taken to include all things connected with such an
analysis. The term ‘critical path analysis’ (CPA) is also used. The analysis
relies on a single duration estimate for each activity.
The networkanalysis takes advantage of the special directedcharacteristics
of the network. In particular, the analysis is done in essentially three stages.
The first stage is a forward pass through the network in which earliest
times (or dates) are calculated for each activity, namely:
Earliest start time (EST)
Earliest finish time (EFT)
52 Conventional treatment from a systems perspective
The second stage is a backward pass through the network in which latest
times (or dates) are calculated for each activity, namely:
Latest start time (LST)
Latest finish time (LFT)
Thereafter (the third stage) the total float (TF), free float (FF) and interfering
float (IF) may be calculated.
F/S relationship calculations
Mathematically, the calculations are as follows for activities with finish-
to-start (F/S) relationships (and zero lead time period) between them. Similar
calculations are performed where other relationships (that is, start-to-start
(S/S), finish-to-finish (F/F), start-to-finish (S/F) and finish-to-start (F/S) – all
with finite lead time periods) apply.
Forward pass
Initial conditions for the first activity(ies):
EST =0
EFT =EST+duration
For all subsequent activities:
EST =maximum EFT of preceding activities
EFT =EST+duration
Backward pass
Terminal conditions for the last activity(ies):
LFT =EFT
LST =LFT−duration
For all previous activities:
LFT =minimum LST of following activities
LST =LFT−duration
Planning tools and terminology A–Z 53
Float
For all activities:
TF =LFT−EFT
=LST−EST
FF =EST of following activity−EFT of activity itself
IF =TF−FF
Setting the earliest start time of the first activity(ies) to zero is optional.
Other values can be prescribed.
For the last activity(ies), setting the latest finish time equal to the earliest
finish time means that the so-called critical activities will have zero total
float. Terminal conditions other than the one mentioned above can be
prescribed and this would be the case, for example, where there was a
desired finish date for the project.
The project completion date is the EFT or LFT of the last activity(ies).
Overlapping relationships
The above calculations can be generalised for all overlapping relationship
types. The notation of Figure 3.7 applies.
A full network analysis requires all network interconnections to be con-
sidered.
The calculations on networks involving overlapping relationships proceed
in a similar fashion to the calculations of elementary network analysis. That
is, on a forward pass, activity EST and EFT information is obtained and
then on a backward pass, activity LFT and LST information is obtained.
The floats and the critical path may then be determined. The relevant
expressions are as follows for a network starting at activity 0 and finishing
at activity N:
Forward pass calculations
EST
0
=0
EFT
0
=EST
0
+DUR
0
EST
j
=max
all t,j









EFT
t
+LT F¡S
EST
t
+LT
t
S¡S
EFT
t
+LT
j
−DUR
j
F¡F
EST
t
+LT
t
+LT
j
−DUR
j
S¡F
54 Conventional treatment from a systems perspective
The above implies that for more than one preceding activity, t, or overlap-
ping dependence relationship choose the maximum EST
j
value because the
longest path through the network is sought. Where EST
j
turns out to be
negative, upgrade this value to 0.
EFT
j
=EST
j
+DUR
j
EST
i
(LST
i
)
EFT
i
(LFT
i
)
EST
i
(LST
i
)
EST
j
(LST
j
)
Activity i
Activity j
LT
EST
j
(LST
j
)
(a)
Activity i
Activity j
LT
i
LT
i
(b)
Activity i
Activity j
LT
j
LT
j
EFT
i
(LFT
i
)
EFT
j
(LFT
j
)
EFT
j
(LFT
j
)
(c)
Activity j
Activity i
(d)
Figure 3.7 Lead time periods; two activities only. (a) Finish-to-start relationship;
(b) Start-to-start relationship; (c) Finish-to-finish relationship; (d) Start-
to-finish relationship.
Planning tools and terminology A–Z 55
Backward pass calculations
LFT
N
=EFT
N
LST
N
=LFT
N
−DUR
N
LFT
t
= min
all t,j









LST
j
−LT F¡S
LST
j
−LT
t
+DUR
t
S¡S
LFT
j
−LT
j
F¡F
LFT
j
−LT
t
−LT
j
+DUR
t
S¡F
The above implies that for more than one following activity, j, or overlap-
ping dependence relationship choose the minimum LFT
t
value.
LST
t
=LFT
t
−DUR
t
Floats
Total float,
TF
t
=LFT
t
−EFT
t
=LST
t
−EST
t
TF
N
will equal zero if the latest finish time is set equal to the earliest finish
time for the last activity N. In such a case the activities with total float
equal to zero will define the critical path. Where the latest finish time is
not set equal to the earliest finish time for the last activity, TF
N
will equal
the difference between these two values and the critical path through the
network will be defined by activities having total floats equal to this number.
Note that where the largest activity LFT is not the network final activity, the
activities at the end of the network (following the activity with the largest
LFT) will be calculated to have a zero total float yet they are not critical
activities. These activities can be allocated float in a post-analysis check.
Free float,
FF
t
= min
all t,j









EST
j
−(EFT
t
+LTi F¡S
EST
j
−(EST
t
+LT
t
i S¡S
EST
j
−(EFT
t
+LT
j
−DUR
j
i F¡F
EST
j
−(EST
t
+LT
t
+LT
j
−DUR
j
i S¡F
The above implies that for more than one following activity, j, or over-
lapping dependence relationship, choose the minimum FF
t
value. These
expressions represent the difference between the right- and left-hand sides
of the earlier expressions for earliest times.
56 Conventional treatment from a systems perspective
A note
For activity-on-node networks with non-overlapping relationships, redun-
dant links occur with triangular logic patterns as shown in Figure 3.8.
Redundancy will similarly occur in networks with overlapping relation-
ships provided the earliest start time of C (Figure 3.8) is determined by
the path A–B–C; redundancy will not occur if the earliest start date is
determined by the link A–C. Hence triangular patterns may or may not
be permissible in networks with overlapping relationships. To check for
redundancy the earliest start time of C has to be computed along the path
A–B–C and along the link A–C. If the value on the link A–C is greater then
remove the redundant link A–C.
Representation
Personal preference dictates the way that the activity start and finish times
and floats are displayed on the network diagram. Figure 3.9 shows a num-
ber of examples. Each person, text and computer package has its own
preference. Computer packages may allow customising the display.
The numerous commercial computer packages available to assist project
planning, all have CPM as their basis.
To many people, network analysis is ‘project management’. The reasoning
goes that if you have a computer package, then you can set yourself up in
business as a ‘project manager’. This is not a realistic appraisal of project
management.
Occasionally people (incorrectly) use the term ‘PERT’ (program evalua-
tion and reviewtechnique) when talking about CPM. PERT is a probabilistic
analysis method and appears not to be used by practitioners (even though
the term PERT is used, but wrongly).
It has been humorously commented that: The critical path method is a
management technique for losing one’s shirt under complete confidence.
A
B
C
Figure 3.8 Redundant link example.
Planning tools and terminology A–Z 57
EST EFT
LST LFT
TF
Duration
Activity
name
Activity name
Duration
EST EFT
LST LFT
TF
(a)
Activity
name
Duration
EST
LST
EFT
LFT
TF
FF
(b)
Duration
EST EFT
TF
Activity name
(c)
Figure 3.9 Example drafting styles.
Cumulative production plot
A cumulative production plot shows an activity’s or project’s production up
to any time or distance (as the independent variable); the slope of the plot
represents the rate of production (Figure 3.10). Figure 3.10 can be either
an activity level or project level representation.
Some variants on Figure 3.10, at the activity level, are:

The axis labelling may be interchanged, with the independent variable
on the vertical axis (Figure 3.11).

The diagram may be drawn down the page (Figure 3.11).

The line plot may be given a finite width where work occupies durations
of significance (Figure 3.12).
For a bar chart such as Figure 3.13a, joining the bars (that is, ‘integrating’
the bar chart) and changing the vertical axis gives the cumulative production
plot of Figure 3.13b.
A cumulative production plot may be referred to as a program or a
schedule, in some situations may alternatively be called a trend plot, trend
graph or time-chainage chart, and is used in the line of balance technique.
Learning
period
Planned rate
of progress/
production
Time or distance
C
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n
(
t
a
s
k
s
,

d
i
s
t
a
n
c
e
,

m
a
t
e
r
i
a
l

u
s
a
g
e
,

.

.

.
)
Figure 3.10 Example cumulative production plot.
Cut
Fill
Subbase
Seal,
surface
Culvert
Subbase
Seal,
surface
Furniture
Clear
Jan
Feb
Mar
Apr
May
Cut
Fill
Culvert
Finished road line
Original ground line
Chainage
Landscaping, fencing
Furniture
Figure 3.11 Example cumulative production plot (time-chainage chart) for road
construction. (Activity names have been abbreviated because of space
limitations – work items are implied.)
Working days
1
2
3
4
5
U
n
i
t
s
P
l
u
m
b
e
r
E
l
e
c
t
r
i
c
i
a
n
R
e
n
d
e
r
e
r
P
l
a
s
t
e
r
e
r
C
a
r
p
e
n
t
e
r
C
a
r
p
e
n
t
e
r
P
a
i
n
t
e
r
Figure 3.12 Example project with sequential/repetitive work. (Activity names have
been abbreviated because of space limitations – work items are implied.)
G
M
1
2
3
4
5
6
7
8
9
P/R
R
Days
I
n
t
e
r
n
a
l

p
a
r
t
i
t
i
o
n
s
F
i
n
i
s
h
e
s
R
o
u
g
h
-
i
n

s
e
r
v
i
c
e
s
a
n
d

w
e
t

a
r
e
a
s
F
l
o
o
r
/
s
t
o
r
e
y
(a)
Figure 3.13 Example construction program for internal work for a building. (a) Bar
chart; (b) Cumulative version. (Activity names have been abbreviated
because of space limitations – work items are implied.)
60 Conventional treatment from a systems perspective
12
10
11
0
1
2
3
4
5
6
7
8
9
Days
(b)
F
l
o
o
r
s
I
n
t
e
r
n
a
l

p
a
r
t
i
t
i
o
n
s
F
i
n
i
s
h
e
s
R
o
u
g
h
-
i
n

s
e
r
v
i
c
e
s
a
n
d

w
e
t

a
r
e
a
s
Figure 3.13 (Continued).
(Refer ‘resource balancing’.) Associated with every cumulative production
plot, a network can be drawn showing the work involved; see ‘resource
balancing’.
When ‘integrated’, a bar chart becomes the same as a time-based cumula-
tive production plot. Cumulative production plots, bar charts and multiple
activity charts contain essentially the same information. It is the additional
vertical scale that makes the plot attractive, compared to a bar chart. The
computational path to the cumulative production plot should be no differ-
ent to that to a conventional bar chart, namely via a network.
In cumulative production plot terminology, a resource schedule is one
determined by resource numbers, while a parallel schedule is one determined
by production.
Cumulative production plots also find use in reporting for replanning
purposes.
Decompression
Decompression refers to the project duration being expanded. The project
decompression problem is a subproblem of the total planning problem,
where emphasis is placed on the different level controls that can lead to
activity and project lengthening, and all other controls are held fixed. As
Planning tools and terminology A–Z 61
with the case of applying additional resource numbers or using higher
resource production rates to decrease (compress) an activity’s duration, so
applying less resource numbers and decreasing the resource production rates
increases (expands) an activity’s duration. An expanded project duration
may be possible if a later completion time (extension) is available or allowed.
Delay
A delay is a time period affecting the undertaking of or impeding the
progress of a project. The causes of delays are many, but only refer to those
where uncertainty is involved.
A delay allowance, delay contingency or ‘time’ reserve may be incorpo-
rated into a project’s schedule where this is possible. In Figure 3.14, o days
have been allowed over the total project. The diagonal line represents a
uniform rate of usage of the delay allowance over the total project.
Figure 3.14 is a project level representation.
Deliverable
The term ‘deliverable’ is loosely used and abused. Its consensus meaning is
that of end-product plus state final (terminal) conditions for the project, or
equivalent for a project milestone. However, because of the term’s abuse,
it is best avoided.
In terms of work breakdown, an end result has no meaning, and so
deliverables should not appear in a work breakdown structure, in spite of
how useful a deliverable might be to some people for project management
purposes.
Refer Carmichael (2004).
Project
start
Project
completion
Time
Total delay allowance
D
a
y
s
q
0
Uniform rate
of delay allowance
usage
Figure 3.14 Delay allowance; cumulative delay plot.
62 Conventional treatment from a systems perspective
Disturbance
The corruption that prevents any planned for project output being obtained
exactly is represented by disturbance. Disturbance leads to the need for
continual monitoring and reporting of project progress, and possibly revised
controls.
Earned value
At any time during a project, earned value is the amount that should have
been spent on the production done.
The earned value of production performed is found by multiplying the
estimated per cent completion for the production by the planned cost for
that production.
Earned value =(Estimated) Per cent complete ×Planned cost
Earned value is reporting at the project level.
Element
An element of an activity represents a portion of an activity equal to the
smallest chosen time unit for the project. In many projects, the smallest
time unit is a day, but other time units are possible. For an activity with a
duration of J, then J activity elements result.
End-product
A project will commonly come about because of an identified need or want
for some product, facility, asset, service and so on. This end-product is
achieved through a project.
This distinction sometimes causes people confusion and many people are
not aware of the distinction, or of the needtomake a distinction. For example,
people sometimes refer to a building as a project. It is not. The processes that
go together to materialise the building are the project. The building is the
end-product. (It is acknowledged that the definition of a project is sufficiently
flexible to include the operation and maintenance phase of a product within
what is called the project. However, this is not the issue here.).
Refer Carmichael (2004).
Error control
Closed loop control may be referred to as error control if the control is
selected based on the difference between desired system behaviour and
actual system behaviour.
Planning tools and terminology A–Z 63
Estimates
One distinction between project costs is that between direct and indirect
costs. Direct costs are costs that are directly proportional to the work quan-
tity. The cost of people, equipment and materials needed to carry out an
activity is an example. The costs discussed in network analysis are generally
direct costs. Indirect costs are one-off and time varying, and accrue whether
or not any work is done. Examples include the salaries of project adminis-
trative staff, hire of accommodation and insurance. The total project cost
is the sum of all the activity direct costs and indirect costs. If developing
a tender price, a contingency might be added to account for uncertainties.
(Also included would be head office overheads and profit [Figure 3.15].)
Estimates for durations and costs to carry out activities are usually based
on past experience (extracted from historical records or ‘time’ studies) and
trends, or could be developed from first principles. The usefulness of esti-
mates depends on the skills of the estimator and the familiarity this person
has with the work involved in the project.
Several degrees of estimate accuracy may be recognised. Generally, the
accuracy of estimates improves as the project stages progress. Early on,
estimates may be quite coarse. At implementation, reasonably accurate
estimates are required.
A common (but not the only) differentiation of degrees of accuracy in
estimates is:

Approximate (within 20–25% accuracy). Usually used to assess the
magnitude of the project and based on duration or cost indices.

Preliminary (within 10–20% accuracy). These are compiled from
overall unit rates for each piece of work in the project, and are usually
used for a decision whether to proceed with the project or not.

Detailed (within 5–10% accuracy). These are used for tendering or
replanning purposes, and they require either accurate unit rate methods
or a bottom-up approach.
Price
Estimated
costs
Contingency
Profit Overheads
Direct
costs
Indirect
costs
Project-based costs
Non-project-based costs
Margin/mark-up
Figure 3.15 Tender price composition.
64 Conventional treatment from a systems perspective
Estimates depend on resource assumptions, local and production factors
including the availability of people and special skills, availability of materi-
als, weather and hazard facts, and availability of equipment/plant.
According to Murphy’s Law (extended): Everything takes longer than
you think. And Nothing is as easy as it looks.
Event
An event is the start or end of an activity; an occurrence at a specific time.
Special or important events may be given names, such as milestone event.
Events correspond to the nodes in activity-on-link diagrams.
Some computer packages may (for computational reasons) represent an
event as an activity of zero duration, but this can lead to misunderstanding;
an event is not an activity.
In terms of work breakdown, an event has no meaning, and so events
should not appear in a work breakdown structure, in spite of how useful
an event is for project management purposes.
Factual network
Also called as-executed (as-built, as-constructed, as-implemented) pro-
grams. A factual network is a network (best done on a timescale) of
as-executed activities indicating changes to the original program, includ-
ing delays and variations, and agreed changes, cross referenced to files on
this information. It is a record of facts. The factual network becomes a
permanent record of a project’s progress. A factual network (as opposed
to the original network) may not lend itself to analysis via conventional
methods. The factual network also finds a use in attempting to resolve dis-
putes involving delays, schedule extensions, liquidated damages and related
claims. It is similar in concept to an as-built drawing, in that it shows the
actual durations to perform the various work items.
Fast-track project
A project is said to be fast-tracked if its phases overlap (Figure 3.16).
A phase is started before the previous phase is complete. Project phases are
run in parallel where they may normally be run sequentially.
At the activity level, a similar situation can occur when start-to-start,
finish-to-finish or start-to-finish relationships are specified on the activity
network, in place of more usual finish-to-start relationships. This can have
the effect of completing the work sooner, which is the intent of fast-tracking.
The primary benefit of fast-tracking is a reduced schedule, and earlier
completion time for the project, that is earlier delivery of the end-product.
The actual project cost could be expected to increase, in return for an earlier
Planning tools and terminology A–Z 65
Phase 1
Phase 2
Phase 3
Phase 4
Phase 1
Phase 2
Phase 3
Phase 4
Conventional project phasing
Fast-tracked phasing
Figure 3.16 Bar chart showing comparison of conventional and fast-tracked
approaches.
return on money invested, for example earlier collection of rental, earlier
sale and decreased cost of borrowed money.
The downside with fast-tracking is the associated peculiar managerial
problems.
Refer Carmichael (2004).
Financial planning
Financial planning refers to all that involved in deciding the usage of money
on a project. The term ‘budgeting’ may also be used to mean financial
planning, and the term ‘budget’ to mean financial plan.
The financial planning problem is a subproblem of the total planning
problem, where emphasis is placed on the different level controls that can
influence money usage (including recoupment), and all other controls are
held fixed. Resource usage is converted to a unit of money.
Budgeting for projects, some might argue, is a more difficult task than
budgeting for a business. Projects are unique and have to be developed from
first principles each time, compared with a business where next year’s oper-
ations could be expected to bear some similarity with this year’s operations.
Though budgeting in a business via work packages (and later monitoring
against these work packages) is little different to that involved in budgeting
for projects.
Budgets tie expenditure to activities. If this is related to an accounting
system, then as work is done, this can be charged against individual account
numbers.
66 Conventional treatment from a systems perspective
All the different participants in a project – owner, contractor, subcon-
tractor, . . . – have money being expended and various forms of income.
Financial planning takes into account expenditure, income and cash flow.
For example, where a contractor (and similarly for a consultant) is being
reimbursed by the project owner for work done, a cumulative income plot
can be superimposed on the cumulative expenditure plot to give Figure 3.17.
Refer ‘S curve’. Incomes and expenditures might be plotted based on dates of
anticipated transfer of funds and billings, when invoiced, when committed,
or when paid, depending on the user.
The comments below apply for both contractors and consultants.
Commonly project work is done by contractors, but not paid for by the
owner until progress claims or similar are submitted. The contractor funds
the project.
The vertical difference between the cumulative outlays of the contractor
and the cumulative reimbursements represents the amount of money the
contractor has to find on any day to finance the project. It may not be until
the end of the project that the contractor shows a profit. Many projects
have a negative cash flow until the very end when final payment, including
retention funds, is received.
There may be large differences between cash flow patterns on different
projects. A positive cash flow is attractive to a contractor since it elimi-
nates borrowing or tying up its own funds (which then become available
for investment elsewhere). Negative cash flow draws on the contractor’s
Project
start
Project
completion
Time
Profit
Initial setting up
of project
Winding down
of project
Progress
payment
Project income
(cumulative)
Project expenditure
(cumulative)
Funds required
to run project
C
u
m
u
l
a
t
i
v
e

m
o
n
e
y
Figure 3.17 Example cumulative expenditure and cumulative income diagrams
(as-planned).
Planning tools and terminology A–Z 67
working capital, or borrowing is necessary. For a project to proceed, funds
availability has to exceed funds requirements.
As shown in Figure 3.17, the contractor is funding the project except for
right at the end of the project. The expenditure exceeds the income, implying
a negative contractor cash flow. Contractors might attempt to improve their
cash flows by moving the cumulative expenditure and cumulative income
diagrams closer together. The ideal from the contractor’s viewpoint is to
have the cumulative income diagram above the cumulative expenditure
diagram, implying a positive cash flow for the contractor; the owner is
funding the project.
The ways that a contractor might improve its cash flow include:

Receiving a down payment, up-front payment, advance payment,
mobilisation payment, part payment on acceptance of tender, or similar
terminology.

Receiving payment of a progress claim in a timely fashion.

Delaying payments to creditors, consultants, subcontractors and sup-
pliers to 30, 60, 90 or more days (a lag between when an invoice is
received and when payment is made).

Unbalancing its bid through front end loading. That is, items that occur
early in a project are bid and reimbursed at marked-up prices/rates.

Starting activities closer to their LST rather than their EST. Planned
expenditure will commonly lie somewhere between these two extremes.
The inflow of funds is also affected by profit gained on the project, and
reworked into the project. Progress payments may be invested and attract
interest.
Awareness of all these issues allows the contractor to put together a
considered bid, or be selective in its bidding on projects.
From the owner’s viewpoint, there may be curiosity in the contractor’s
cash flow, but possibly more curiosity in the owner’s payment schedule
which is the contractor’s income plot. Owners may try to defer payment to
a contractor for as long as it does not hinder the contractor’s performance.
Before awarding the work to a contractor, the owner may need to have
an idea as to what its total costs for the project will be, and when it will be
required to make progress payments (and hence by when money should be
obtained).
Feasibility, from an owner’s perspective, would generally be established
based on a master plan level of generality.
The cost of financing
On large projects, the cost of financing can be a major consideration.
Interest on borrowed money can represent a significant contribution to the
68 Conventional treatment from a systems perspective
project cost. Therefore, expenditure tends to be postponed for as long as
possible, subject to not having a deleterious effect on a project’s progress.
On the other hand, escalation costs are hard to predict. (Commonly, an
escalation allowance is calculated for the midpoint of the time span over
which the project work is expected to be performed. Escalation is assigned
to each work package relative to the duration of each package.) The ear-
lier a project is completed, the cheaper will be the associated end-product,
and end-product income can start to be received. To guard against esca-
lation, components may be purchased early, but this introduces inventory
costs, which include deterioration, damage and even theft while in storage.
Cash flow calculations accordingly should include interest, escalation and
inventory costs.
Consideration of all these issues will lead to a minimum cost solution.
Float
Also called slack. The term ‘float’ is used in this book. Float is a free time
period. Float is the time period available to delay the start or finish of an
activity without affecting the project duration. Conventionally, non-critical
activities have float; critical activities have no float.
Several types of float may be distinguished. The most popular are total
float (TF), free float (FF) and interfering float (IF). They are related through,
TF =FF+IF
The definitions of all floats are similar. All are time periods available to
delay the start or finish of an activity without affecting the project duration.
Total float is the total time period available. It may be thought of as being
composed of free float and/or interfering float. Free float is a time period
available, but if used will not affect the start times of following activities.
Interfering float is a time period available, but if used will affect (interfere
with) the start times of following activities.
Whether float is classified as free float or interfering float depends on
where the activity occurs in the network. For example, consider the part
activity-on-node network of Figure 3.18. The critical path runs through the
bottom path and has a duration of 12 days. The top path has a combined
duration of 3+2+4 =9 days, giving a total float of 12−9 =3 days. This
3 days appears as interfering float in activities A and B, but as free float
in activity C. It is the same float manifesting itself in different ways, and
may be used in activities A, B and C in any combination, but only up to
a combined total of 3 days. It is called interfering float in activities A and
B, because if used in A and/or B, it affects the start time of C. It is called
free float in C, because if used in C, it does not affect the start time of any
other activity.
Planning tools and terminology A–Z 69
A
3
B
2
C
4
D
12
TF = 3
FF = 0
IF = 3
TF = 3
FF = 0
IF = 3
TF = 3
FF = 3
IF = 0
TF = 0
FF = 0
IF = 0
Figure 3.18 Part-network illustrating float type. Activities are indicated by the letters
A, B, C and D, and their durations are given within the circles.
Other floats are mentioned by various writers, for example independent
float and scheduled float.
Where the float is negative, this implies that the required project com-
pletion date is sooner than the project completion date obtained through
a critical path analysis. Activity durations will have to be reduced by this
float amount for the required completion date and calculated completion
date to be the same.
Gantt chart
Refer ‘bar chart’.
Human resource planning
Human resource planning refers to all that involved in deciding the usage
of people on a project.
The human resource planning problem is a subproblem of the total plan-
ning problem, where emphasis is placed on the different level controls that
can influence the usage of people resources, and all other controls are held
fixed.
Lead time period
Also called lead time and lag. It is denoted LT in this book. Refer ‘overlap-
ping relationships’. Lead time period refers to the duration between start-
ing/finishing a preceding activity and starting/finishing a following activity.
70 Conventional treatment from a systems perspective
Lead time period may be expressed as an absolute value, in days, for
example, or as a percentage of the duration of an activity. When expressed as
a percentage, then reducing the activity duration will reduce the associated
lead time period; this situation, of course, does not happen where lead time
periods are expressed as absolute values.
Note that some writers make a distinction between ‘lead’ and ‘lag’. That
distinction is not followed in this book.
Learning
As an individual, team or organisation gains experience in carrying out its
allotted work, the effectiveness of the individual etc. increases. There is a
saying ‘practice makes perfect’. Typically the time period, for example, to
make or do a first unit is higher than the time period to make or do a second
unit and this time period continues to decline for further units though to a
slightly lesser extent.
Generally, as a task is repeated, performance improves with a tapering
off after carrying out the task many times. Knowledge of such possible
improvements in performance can prompt a redesign of the work method;
the resultant time period savings may be incorporated into the plan or kept
as a reserve.
Figure 3.19 shows this schematically and is called a learning curve. The
curve is applicable for individuals, teams and organisations and has been
substantiated in many field observations. Such information has direct rel-
evance to planning any repetitive process, or converting processes into
repetitive ones.
In bar chart form, the learning effect might be represented as in
Figure 3.20.
Units
W
o
r
k
/
u
n
i
t
Figure 3.19 Learning curve.
Planning tools and terminology A–Z 71
Unit 1
Unit 2
Unit 3
etc
Non-learning
Learning
Figure 3.20 Bar chart representation of learning effect. The bars refer to the work
involved with each unit.
The learning curve is commonly described by an equation of the form,
T
n
= T
1
n
b
where
T
n
average work/unit after n units
T
1
work/unit for the first unit
b index of learning (b - 0i
When plotted on logarithmic coordinates, this exponential curve becomes
a straight line (Figure 3.21).
Different values of b, the index of learning, represent different values of
the rate of learning as shown in Figure 3.22.
Conventionally, however, the rate of learning is specified as a percentage
such as 70%, 80%, 90%, . . . For example, an 80% curve means that
the work involved with the nth unit is 80% of that required for the
ln(unit)
ln(n)
T
1
I
n

(
w
o
r
k
/
u
n
i
t
)
I
n

(
T
n
)
Figure 3.21 Learning curve on logarithmic scales.
72 Conventional treatment from a systems perspective
n
b = – 0.1
b = – 0.3
b = – 0.5
T
1
T
n
Figure 3.22 Family of learning curves.
(n−1ith unit. 70%, 80% and 90% curves correspond to b values of about
−0.51, −0.32 and −0.19 respectively.
Both T
1
and b require estimating. Typically, these are adapted from
similar situations or historical records.
Level
A project may be decomposed to subprojects, to sub-subprojects, . . . , to
activities, elements and constituents. Levels refer to the strata in such a
hierarchical decomposition.
Line of balance technique
The line of balance technique applies to resourcing on projects that typically
have sequential/repetitive work content (for example, the work of trades in
constructing a high-rise building). It is a process-oriented approach which
has its origins in the manufacturing industry. Refer ‘resource balancing’,
and ‘cumulative production plot’.
Milestone
A milestone is an important event occurring in a project. A project may have
a number of milestones. Milestones are shown on a network or bar chart
in order that everyone is aware when particular important events occur, or
something should be accomplished by.
As a computational and representational device, some computer packages
may define a milestone by setting a bogus activity’s duration to zero, but
nevertheless a milestone is an event not an activity.
Planning tools and terminology A–Z 73
Monitoring
Monitoring involves measuring or observing the output or performance of
a project.
Monte Carlo simulation
Monte Carlo simulation, as used in planning, is a numerical probabilistic
method of network analysis. However, note that Monte Carlo simulation
is a technique for analysing nearly anything that contains probabilities.
The currently popular way to realistically incorporate variability in activ-
ity durations (typically represented by probability distributions) for network
analysis is to use Monte Carlo simulation.
Monte Carlo simulation essentially converts a difficult probabilistic prob-
lem into many simpler deterministic problems. The steps involved for net-
work analysis are:

Activity durations are sampled (via the generation of uniform random
numbers).

A conventional network analysis (incorporating overlapping relation-
ships if applicable) is carried out.

The above two stages are repeated many times.

Relevant statistics on important items, such as project completion times,
are collected.
The first step is one of data generation, the second step is analysis, while
the fourth step is bookkeeping.
Monte Carlo simulation is suited ideally to computer use, where large
amounts of data can be handled with ease. It is not a technique for use by
hand or where a closed-form solution is required.
Some commercially available Monte Carlo simulation packages sit as
overlays on critical path method (CPM) packages.
Multiple activity chart
Also called a multi-activity chart, it is a chart on which the activities of
more than one resource (people or equipment) are recorded on a common
timescale to show their interrelationship.
A multiple activity chart is drawn in a similar sense to bar charts with
a timescale and shading length in proportion to the duration of whatever
is being plotted. Separate bars are drawn for each resource. The bars may
be drawn vertically (top to bottom time sense) or horizontally (left to right
time sense).
74 Conventional treatment from a systems perspective
Time
Machine Operator
Supervisor
Not involved in work
Involved in work
Figure 3.23 Example multiple activity chart.
The chart shows very readily periods of idleness and periods when the
resource types are utilised.
Figure 3.23 shows an example of multiple activity chart. Times for such
charts are typically based on field records.
A multiple activity chart is essentially a different form of bar chart con-
centrating on what particular resource types are doing, usually without
connections, but there is no reason why connections cannot be included.
Bar charts tend to be developed beforehand, while multiple activity charts
tend to show what has happened; apart from that, they are no different.
Cumulative production plots, bar charts and multiple activity charts con-
tain essentially the same information.
Murphy’s Law
The origin of Murphy’s Law, the original principle ‘If anything can go
wrong it will ’ is unclear. Some of the published attribution is not convincing.
Possibly the naming came about because Murphy is a distinctly Irish name,
Planning tools and terminology A–Z 75
and the Irish are known for their quirky humour. Many of the following
examples are from Murphy’s Law, A. Bloch, Methuen, London, 1986.
Nagler’s comment on the origin of Murphy’s Law
Murphy’s Law was not propounded by Murphy, but by another man of the
same name.
The reliability principle
The difference between the Laws of Nature and Murphy’s Law is that with
the Laws of Nature you can count on things screwing up the same way
every time.
O’Toole’s commentary on Murphy’s Law
Murphy was an optimist.
Goldberg’s commentary
O’Toole was an optimist.
There are many extensions and corollaries of the basic principle behind
Murphy’s Law, and again the authorship of these is unclear. All give a
humorous view of laws guiding life.
Murphy has something to say on planning related matters. Some exam-
ples are:
No matter what goes wrong, it will probably look right.
Nothing is as easy as it looks.
Everything takes longer than you think.
Before you can do something, you have to do something else.
The nearer to completing the job, the greater the alterations required.
Now they tell us Law
Information calling for a change in plans will arrive after the plans
are complete.
Sodd’s First Law
When a person attempts a task, he or she will be thwarted in that task
by the unconscious intervention of some other presence (animate or
inanimate). Nevertheless, some tasks are completed, since the inter-
vening presence is itself attempting a task and is, of course, subject to
interference.
Pudder’s Law
Anything that begins well, ends badly.
Anything that begins badly, ends worse.
76 Conventional treatment from a systems perspective
Wynne’s Law
Negative slack tends to increase.
3rd Corollary to Law of Applied Confusion
After adding two weeks to the schedule for unexpected delays, add
two more for the unexpected, unexpected delays.
Laws of Computerdom according to Golub
1. Fuzzy project [goals] are used to avoid the embarrassment of
estimating the corresponding costs.
2. The carelessly planned project takes three times longer to complete
than expected; a carefully planned project takes only twice as
long.
3. The effort required to correct course increases geometrically with
time.
4. Project teams detest weekly progress reporting because it so
vividly manifests their lack of progress.
Patton’s Law
A good plan today is better than a perfect plan tomorrow.
Frothingham’s Fallacy
Time is money.
Westheimer’s Rule
To estimate the time it takes to do a task: estimate the time you think
it should take, multiply by 2, and change the unit of measure to the
next highest unit. Thus we allocate 2 days for a one-hour task.
Ninety–Ninety Rule of Project Schedules
The first ninety percent of the task takes ninety percent of the time,
and the last ten percent takes the other ninety percent.
Cheop’s Law
Nothing ever gets built on schedule or within budget.
The Einstein Extension of Parkinson’s Law
A work project expands to fill the space available.
Corollary. No matter how large the work space, if two projects must
be done at the same time they will require the use of the same part of
the work space.
Workshop Principle (number 4)
The more carefully you plan a project, the more confusion there is
when something goes wrong.
Planning tools and terminology A–Z 77
McDonald’s Corollary to Murphy’s Law
In any given set of circumstances, the proper course of action is
determined by subsequent events.
Seay’s Law
Nothing ever comes out as planned.
Sweeney’s Law
The length of a progress report is inversely proportional to the amount
of progress.
First Law of Corporate Planning
Anything that can be changed will be changed until there is no time
left to change anything.
Kent Family Law
Never change your plans because of the weather.
The following Murphy’s Calendar is useful for planners.
FRI MON TUES WED THURS MON SAME DAY
7 6 5 4 3 2 1
14 13 12 11 10 9 8
21 20 19 18 17 16 15
28 27 26 25 24 23 22
35 34 33 32 31 30 29
This calendar has certain advantages:

Every job is a rush job. Everyone wants things done yesterday. With
this calendar an order given on the 7th can be carried out on the
3rd.

Everybody wants things done early – on Mondays for preference. So
there are two Mondays in each week.

There are several extra days at the end of the month for those end of
the month rushes.

There are no bothersome non-productive Saturdays and Sundays.

There is a new day each week – ‘same day’. On this day, ‘while you
wait’ and ‘same day’ jobs may be handled without interruption to other
promises. Everyone will be happy and ulcer free.

Acceptance of this calendar is made easier by pointing out that this
system has been in unofficial use in many companies for some years.
78 Conventional treatment from a systems perspective
Network
A network is a diagram showing the logical sequence of undertaking activi-
ties. In other words, it shows the precedence relationships between activities,
the work order interrelationship of the activities, or the logic of how the
project is to be undertaken. In drawing networks, care has to be exercised
that false logic is not introduced.
Networks are the basis of project planning and, as part of an iterative
analysis approach to planning, represent a rational intermediary step to
project programs, resource and expenditure plots and S curves.
In general, a network is composed of links and nodes. For networks used
in planning, the links are directional and are represented by arrows; the
network may accordingly be called a directed network. This gives rise to
the two fundamental types of networks:

Activity-on-link diagrams (elsewhere called arrow diagrams or activity-
on-arrow diagrams) whereby activities are represented by arrows start-
ing and finishing at nodes (also called starting and finishing events)
(Figures 3.24a and 3.25a).

Activity-on-node diagrams (elsewhere called precedence diagrams or
circle diagrams) whereby activities are represented by the network
nodes, with arrows linking the activities (Figures 3.24b and 3.25b).
Terminology used in this book is activity-on-link diagram and activity-on-
node diagram, and both are collectively referred to as precedence diagrams,
even though this is not the practice in other publications which generally
use conflicting terminology.
Project
activity
(a)
Project
activity
(b)
Figure 3.24 Network diagram types. (a) Activity-on-link diagram; (b) Activity-
on-node diagram.
Planning tools and terminology A–Z 79
Assemble
options
Apply
finishings
Marketing
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Dummy
Dummy
(a)
Assemble
options
Apply
finishings
Marketing
(b)
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Figure 3.25 Example networks. See the associated bar chart, Figure 3.1, and time-scaled
network, Figure 3.55. (a) Activity-on-link diagram; (b) Activity-on-node dia-
gram. (Activity names have been abbreviated because of space limitations –
work items are implied.)
The same analysis is used for both network types, and the two types of
diagrams give the same answers. It is possible to develop a transformation
between the two types of diagrams.
It is a matter of personal preference which network type it used, although
the orientation of computer packages accessible to the user is an influence.
Clients may also insist on the use of a particular computer package which
in turn will often determine the type of network used. Some computer
packages allow both types of diagrams. Activity-on-link diagrams were the
basis of early critical path analysis work in the 1960s and 1970s. Recent
favour by users seems to be towards activity-on-node diagrams.
Some people insist that activity-on-node diagrams are much better than
activity-on-link diagrams, yet the same people use connected bar charts
to convey a project’s program; that is, both activity-on-node and activity-
on-link diagrams are being used simultaneously. Many unknowingly use
activity-on-link diagrams (to a timescale) when using certain software pack-
ages; the activity-on-link diagram might be disguised as a connected bar
chart – this is in fact a time-scaled activity-on-link diagram.
In favour of activity-on-link diagrams, bar charts are universally used,
and activity-on-link diagrams are the only type that can be represented to
a timescale.
80 Conventional treatment from a systems perspective
In favour of activity-on-node diagrams, they allow easier modification
(editing or updating) once initially developed. Adding and taking away
links in activity-on-node diagrams tends to be easier than adding and taking
away nodes in an activity-on-link diagram. The activity-on-node diagram
tends to be easier to understand and follow by all strata of management and
permits an easier numbering or coding scheme. An activity can be connected
to other activities directly; with an activity-on-link diagram there may be a
need to introduce a dummy activity in order to not affect the required logic.
For overlapping relationships, such relationships apply between activities
and these are easier to represent on an activity-on-node diagram.
Trade A Trade B Trade C etc.
Storey 1
Storey 2
Storey 3
etc
(a)
Trade A Trade B Trade C etc.
Storey 1
Storey 2
Storey 3
etc
A
A
B
B C
C
Dummies
Dummies
Dummies
(b)
Figure 3.26 Part activity-on-node and activity-on-link diagrams applying to the construc-
tion of a multistorey building. (a) Activity-on-node diagram; (b) Activity-
on-link diagram; the vertical arrows between storeys are dummy activities.
(Activity names have been abbreviated because of space limitations – work
items are implied.)
Planning tools and terminology A–Z 81
There is a special case where activity-on-node diagrams are preferred to
activity-on-link diagrams, and that is where there are the same sequential
activities repeating. See for example, Figure 3.26, which relates to the con-
struction of a multistorey building; each trade in sequence works its way up
the building. The activity-on-node diagram turns out to be a more econom-
ical representation compared to the activity-on-link diagram, which needs
the incorporation of many dummy activities to maintain the logic.
For other ‘linear’ type projects, for example relating to roads, pipelines,
railway lines and like structures involving sequential/repetitive activities,
activity-on-node diagrams may also be preferred.
In drawing a network:

The diagram is not drawn to any scale.

The diagram has a left to right sense, where the project start is at the
left, and project end is at the right. Activities to the left of the diagram
could be expected to occur before activities to the right of the diagram.

Arrows do not have to be straight. The length and direction of the
arrows is not important, but generally have a left to right sense. Where
arrows cross over, pipeline style drafting can be used to help reduce
possible confusion.

Activities can be located anywhere vertically on the page. It may be
found convenient though to locate activities associated with a certain
trade or function together, and also to locate the important activities
near the (vertical) centre of the page.

Various drafting styles are possible for nodes, ranging from circles
to multicelled rectangles. The size and form of the shapes are not
important. Figure 3.9 shows some examples.

A following activity may depend on a preceding activity(ies) according
to any of four overlapping relationships – finish-to-start (F/S), start-to-
start (S/S), finish-to-finish (F/F) or start-to-finish (S/F). Multiple rela-
tionships may also occur between activities.

Activities that can be done concurrently are drawn ‘in parallel’.

Activities that are dependent are drawn ‘in sequence’.

Multiple dependency gives rise to multiple links into and/or out of an
activity (activity-on-node diagram) or multiple activities into or out of
nodes (activity-on-link diagram).

Dummy ‘Project Start’ and ‘Project Finish’ activities can be added for
completeness, but are not essential.

Circular logic or circuits are not possible. Redundant logic can be
removed.

Where numbering (optional) is used for links and/or nodes, the num-
bering scheme can usually be anything the user chooses, although a
computer package may require a specific convention.
Refer Carmichael (1989).
82 Conventional treatment from a systems perspective
Network analysis
Refer ‘critical path method, CPM’ (equivalently critical path analysis, CPA),
‘PERT’ and ‘Monte Carlo simulation’.
Objectives
The plural terms ‘objectives’ and ‘constraints’ are generally used in this
book even though the singular forms may be applicable in some situations;
the exception is where the singular is deliberately intended.
The materialisation of a project’s end-product can be performed, possibly,
in an infinite number of ways. In systems studies terms, the problem is an
inverse problem, and hence there is no unique solution.
The criteria by which the preferred materialisation of the end-product
is selected are the project objectives. Work method, scope, resource and
resource production rate considerations on projects follow. The selection
of the preferred materialisation or means to the end-product is the solution
to the planning problem.
In a similar idea, there are possibly an infinite number of versions of
end-products. The criteria by which the preferred end-product is selected
are the end-product objectives. Form, function, finishes etc. of the end-
product follow. The selection of the preferred end-product form is a design
problem.
That is, on any project there are two types of objectives:
1. End-product objectives
2. Project objectives (Figure 3.27).
Commonly, project objectives say something about project cost, project
duration and deviation from specification, but other objectives are possi-
ble. And these may apply throughout the project (for example, minimum
Objectives
Hierarchical levels of objectives
End-product
objectives
Project
objectives
Figure 3.27 End-product and project objectives.
Planning tools and terminology A–Z 83
deviation from specification), or at the final (terminal) point of the project
(for example, minimum duration which is related to the project completion
time). That is, general project objectives will contain a component over
the time domain of the project and a component at the final (terminal)
point.
Objectives may be expressed at the various project, activity, element and
constituent levels.
End-product and project objectives may derive from higher level values
within an organisation, for example originating from a corporate plan. Such
values may also reflect political, marketing, environmental (natural), . . .
concerns.
End-product objectives and project objectives (and constraints) may
relate, for example, to:

Money (end-product – sales, benefit:cost ratio (BCR), net present value
(NPV), . . . ; project – cost, budget, . . . )

Duration (end-product – lifetime, . . .
˙
; project – duration, . . . )

Resource usage

Quality issues

Community acceptance

Environmental (natural) effects

Safety

Risks

Public impact

Extreme event impact (floods, cyclones, . . . )

Social impacts

Geotechnical considerations.
These are expressed in terms of end-product matters or project matters, as
the case may be.
Refer Carmichael (2004).
Alternative names for objective. As needed to perform systematic problem
solving/systems synthesis, alternative names include:

(Optimality) criterion

Performance index

Performance measure

Merit function

Payoff function

Figure of merit

Aim
84 Conventional treatment from a systems perspective

Goal

Cost function

Design index

Target function.
These generally occur outside the management literature. They are com-
mon in the optimisation, optimal design and optimal control literature
(Carmichael, 1981).
Open loop control
In open loop control, the control is selected up front, and no follow-up
monitoring of system performance and corrective control or action is car-
ried out.
Output
The output describes the external or observable system response, perfor-
mance or behaviour. For the stripped-back projects considered in this book,
output and state are the same, that is there are no observability (in the sense
of Kalman) issues. Output contains multiple parts (that is, it is a vector
quantity of output variables). Output is a controlled variable.
Overlapping relationships
An elementary development of networks assumes that when one activity
finishes another can start (with zero lead time period between the activities).
The activities are said to have finish-to-start (F/S) relationships. Generalisa-
tions of this are possible to embrace various relationships between the starts
and finishes of dependent activities, namely start-to-start (S/S), finish-to-
finish (F/F), start-to-finish (S/F), and finish-to-start (F/S) relationships with
non-zero lead time periods (LT) (Figure 3.28). For example, a start-to-start
relationship implies the following activity cannot start until some specified
time period after the previous activity has started, while a finish-to-finish
relationship implies the following activity cannot finish until some specified
time period after the previous activity has finished.
Planners seem to have most usage for F/S, S/S, F/F and S/F relationships
in that order.
Using S/S, F/F and S/F activity overlapping relationships can be like
fast-tracking at the activity level; activities run in parallel instead of
sequentially.
Planning tools and terminology A–Z 85
LT
i
LT
j
LT
Finish-to-start (F/S) with lead time period
Start-to-start (S/S) with lead time period
Finish-to-finish (F/F) with lead time period
Start-to-finish (S/F) with lead time period
LT
j
LT
i
Figure 3.28 Possible overlapping relationships between activities; bar chart
representation.
Subactivities
Many people find the overlapping facility helpful, and also economical in
terms of representations such as the network diagram, bar chart and so
on. However, overlapping relationships can be avoided through the use of
subactivities. See for example, Figures 3.29–3.31.
Example
A situation in which start-to-start relationships, for example, find appli-
cation is illustrated in Figure 3.32, which relates to the construction of a
86 Conventional treatment from a systems perspective
Time
i
j
(a)
LT
i
=

3 days
First 3 days Remainder
LT
i
= 3 days
LT
i
= 3 days
(b)
(c)
(d)
Following trade (Ftr)
Starting trade (Str)
Part
Str
Ftr
Str Ftr
Str
Ftr
Part
Str
Figure 3.29 Start-to-start relationship example. (a) Bar chart; (b) Without using
overlapping ideas; (c) Using overlapping notation; (d) Alternative draft-
ing style. (Activity names have been abbreviated because of space
limitations – work items are implied.)
multistorey building; each trade works its way up the building. Start-to-start
relationships may apply between trades in order to accelerate the comple-
tion of the building; it may not be necessary for one trade to completely
finish on a floor/storey before the following trade commences.
For other ‘linear’ type projects, for example, relating to roads, pipelines,
railway lines and like structures, start-to-start relationships may also be
preferred.
Planning tools and terminology A–Z 87
Time
i
j
(a)
(c)
(b)
(d)
Finishing
treatment (Ftr)
Preparation (Prep)
First part
Second part
Part
Ftr
Part
Ftr
Prep
Prep Ftr
Prep
Ftr
LT
j
=

2 days
LT
j
=

2 days
LT
j
=

2 days
Figure 3.30 Finish-to-finish relationship example. (a) Bar chart; (b) Without overlap-
ping ideas; (c) Using overlapping notation; (d) Alternative drafting style.
(Activity names have been abbreviated because of space limitations –
work items are implied.)
Example
A situation in which start-to-finish relationships, for example, find appli-
cation is illustrated in Figure 3.33, which relates to imposing a completion
time limit dependent on the start time of the first activity. Figure 3.33a
shows the original network. Figure 3.33b introduces a new S/F relationship
between the first and last activity. For example, the start of the first activity
88 Conventional treatment from a systems perspective
Const./
replant
Remove
trees
Time
i
j
(a)
First 3 days
Part
remove
Part
remove
Const.
Replant
trees
Remove
trees
Const./
replant
Remainder
Remove
trees etc
Construction and
replant trees
(b)
(c)
(d)
LT
j
=

2 days
LT
i
=

3 days
LT
i
=

3 days
LT
j
=

2 days
LT
i
+

LT
j
=

3

+

2

=

5 days
Figure 3.31 Start-to-finish relationship example. (a) Bar chart; (b) Without overlap-
ping ideas; (c) Using overlapping notation; (d) Alternative drafting style.
(Activity names have been abbreviated because of space limitations –
work items are implied.)
and the end of the last activity may represent milestones constrained by a
given time interval between when they occur.
Example
In constructing a building, each floor follows the next. For example,
Figure 3.34 shows part of the construction in bar chart form.
Planning tools and terminology A–Z 89
Trade A Trade B Trade C Trade D
Storey 1
Storey 2
Storey 3
Storey 4
S/S
S/S
S/S S/S
S/S
S/S
S/S
S/S S/S
S/S
S/S S/S
S/S
S/S
S/S S/S
F/S
F/S F/S F/S
F/S
F/S
F/S
F/S F/S
F/S
F/S F/S F/S
F/S F/S F/S
Figure 3.32 Part activity-on-node diagram applying to the construction of a multi-
storey building. (Work items implied.)
Act.
1
Act.
1
Act.
2
Act.
2
Act.
3
Act.
3
Act.
4
Act.
4
(a)
(b)
Act. 1
Act. 2
Act. 3
Act. 4
F/S
S/F
F/S F/S
F/S F/S F/S
F/S
F/S
F/S
S/F
Figure 3.33 Example usage of start-to-finish relationship.
In Figure 3.34a, conventional finish-to-start (F/S) relationships are
used.
Figure 3.34b is sometimes termed a ladder representation. The second
activity has float at its start because the F/F relationship with the first
activity dominates. The S/S relationship dominates between the second and
third activities.
90 Conventional treatment from a systems perspective
F/F lead time
period 10 days
F/F lead time
period 3 days
S/S lead time
period 3 days
S/S lead time
period 10 days
(b)
(a)
Ground floor slab
Columns
First floor slab
40 days
12 days
40 days
0 40 52 92 Days
Ground floor slab
Columns 12 days
First floor slab 40 days
40 days
0 Days 31 40 43 74 34
Figure 3.34 Example, building construction. (a) Conventional F/S relationships; (b)
Using S/S and F/F relationships. ( Work items implied.)
Pareto rule/principle
The Pareto rule/principle is named after Vilfredo Pareto (19th century). It
is sometimes called the 80:20 rule/principle. Originally applied to wealth
distribution, it has since been generalised: 80% of the outcomes (outputs)
are the result of 20% of the influences (inputs). The numbers 80 and 20
are not to be interpreted precisely.
There are many applications in project management. For example, 80%
of a project cost estimate is the result of 20% of the component cost items;
80% of a person’s output is due to 20% of that person’s tasks.
Parkinson’s Law
Parkinson’s (first) law
Work expands so as to fill the time available for its completion (Parkin-
son’s Law or the Pursuit of Progress, C. Northcote Parkinson, John Mur-
ray, London, 1957). The thing to be done swells in perceived importance
Planning tools and terminology A–Z 91
and complexity in a direct ratio with the time period to be spent in its
completion.
Parkinson’s (second) law
Expenditures rise to meet income.
Parkinson’s law of delay
Delay is the deadliest form of denial.
Performance measure
Performance measures are gauges by which the success or otherwise of a
project is measured. A performance measure is in the lay usage sense (but
not in an optimal systems sense) of the word ‘objective’ (mentioned in
Carmichael, 2004).
Performance measures tend to be developed later in projects (after the
initial planning has been done); objectives are developed very early. A
performance measure could be expected to follow planning baselines or
targets (as-planned values) for resource usage (and cost and production). A
performance measure may also be based on some industry, competitor or
previous project benchmark.
A performance measure, from any of its sources, could be used as the
basis for error control thinking.
A performance measure might be termed a key performance indicator
(KPI) or similar, but all of these types of terms are used very loosely by
most people. The terms are used frequently by project personnel because
they sound good and impress, but lack precision.
PERT
PERT (acronym for program evaluation and review technique) is a proba-
bilistic first-order method of network analysis where the duration of each
activity is described by a probability distribution derived from an optimistic
estimate, a pessimistic estimate and a most likely estimate (Figure 3.35).
The associated calculations enable the probabilities of completing events
by certain times or keeping to schedule to be calculated. They allow the
ranking of activities (an indication of their criticality) according to the prob-
ability that the free time period at events is greater than or equal to zero.
This information may be used to schedule resources so as to reduce the
possibility of delays within a project.
The analysis by PERT is very similar to that involved in the critical path
method (CPM – a deterministic method where the durations of activities are
estimated to single values) calculations except that PERT carries along one
92 Conventional treatment from a systems perspective
Optimistic
duration
Pessimistic
duration
Most likely
duration
Activity duration
F
r
e
q
u
e
n
c
y

o
f

o
c
c
u
r
r
e
n
c
e
Figure 3.35 Probability density function for activity duration.
additional quantity in the calculations, namely the activity duration vari-
ance. That is, PERT calculations work with an activity expected duration
and a measure of the scatter or variability in the duration (the variance).
(Variance is used here in the probabilistic sense.)
No one appears to be using PERT in practice because such probabilistic
approaches are not popular. Probabilistic network analyses appear to be
best handled using the technique of Monte Carlo simulation. PERT appears
in nearly every textbook on project planning because the technique is useful
as a teaching tool to make planners more aware of the assumptions on which
deterministic network calculations are based. Planning is an inherently
probabilistic problem although in practice it may not be treated as such.
PERT uses an activity-on-link network although events may be given
descriptions. It is this characteristic, of distinguishing events along with
concentrating the calculations on event times, that can give the misleading
impression that PERT uses an activity-on-node diagram as its basis.
Many people and some computer packages use the term‘PERT’ but incor-
rectly–theyare infact meaningCPM. There is muchabuse of the term‘PERT’.
CPM might be considered a special (deterministic) case of PERT with a
couple of differences. PERT is only developed on activity-on-link networks,
and compressing networks in a probabilistic sense raises its own issues.
When the deterministic calculations of the critical path method are car-
ried out, a single project completion time is obtained. To achieve lesser
completion times, the project requires compressing. However, PERT gives
not only an expected project completion time but also the scatter that may
be anticipated in the project completion time. It is thus possible that earlier
completion times could be obtained (without the need for compression)
if activity durations occur closer to their optimistic durations rather than
Planning tools and terminology A–Z 93
Crashed activity (c)
Compressed activity (p)
Normal activity (n)
Time
m
c
t
m
p
t
m
n
t
F
r
e
q
u
e
n
c
y

o
f

o
c
c
u
r
r
e
n
c
e
Figure 3.36 Crashing an activity showing mean times m
t
.
their most likely durations. A similar argument applies to completion times
greater than the expected project completion time.
When network compression is required, it may be carried out similarly
to the deterministic case. Corresponding to each expected project comple-
tion time there is a variance. Although earlier project completion times
may have a finite probability, this probability gets smaller as the project
completion time gets earlier. To ensure a higher probability of achieving an
earlier project completion time, some compression must be carried out. The
situation in Figure 3.36 indicates the same situation for a single activity but
the concept is applicable for both activity durations and project durations.
Cost slopes may be defined in terms of activity expected durations and
associated costs. In general, network compression using PERT is best used
as a guide as to where to apply attention in the network rather than giving
absolute results.
Plan, planning
Planning establishes how and what work will be carried out, in what order
and when and with what resources (type, and number or quantity, addi-
tionally expressed in a money unit). To plan (verb) is the act of choosing
the controls (method, resources and resource production rates) throughout
the project duration. A plan (noun) is the outcome of planning.
Precedence
Precedence refers to the logical connection or dependence between activities
as indicated by a chosen work method. The dependence may be to the start
or finish of a whole or part activity.
94 Conventional treatment from a systems perspective
Precedence may be established, amongst other ways, by addressing the
following questions for each activity, in whole or in part:
Q1. Which activities does this activity follow or logically depend upon?
Q2. Which activities does this activity logically precede?
Q3. Which activities can be done at the same time as this activity?
Answers to these questions establish the precedence relationships between
activities, and the logic of how the project is to be put together.
The resulting network diagram is no more than the answers to Q1, Q2
and Q3, and represents the logic of how the project is to be carried out.
Generally, if one (B) activity’s progress is dependent on another (A)
activity’s progress, then the network diagram will have activity B connected
from activity A. Refer ‘network’ and ‘overlapping relationships’.
Process charts
Process charts portray a sequence of activities diagrammatically by means
of a set of process chart symbols to help a person visualise a process as a
means to examining and improving it (Currie, 1959; ILO, 1969).
An outline process chart gives an overall picture by recording in sequence
only the main operations and inspections.
A flow process chart sets out the sequence of the flow of a product or a
procedure by recording all activities under review using appropriate process
chart symbols. A resource type flow process chart records how the resource
is used.
Process charts are constructed using standard symbols defined in
Figure 3.37. The work is broken down into its component activities which
are represented schematically by these symbols. Further breaking down of
the work may be possible and is sometimes done.
Symbol Activity Predominant result
OPERATION
Produces, accomplishes, furthers
the process
TRANSPORT Travels
STORAGE Holds, keeps or retains
D
DELAY Interferes or delays
INSPECTION Verifies quantity and/or quality

Figure 3.37 Symbols for process chart construction (Currie, 1959; ILO, 1969).
Planning tools and terminology A–Z 95
An outline process chart is useful in design and planning. A method study
is carried out before any work begins. The chart gives an overall view of
the work and graphically shows the sequence of activities where inspections
occur and where extraneous materials, information etc. are introduced.
It does not show who carries out the work or where the work is car-
ried out. Accordingly, only the OPERATION and INSPECTION symbols
are used.
Figure 3.38 shows an example outline process chart. Activities can be
numbered in order, and descriptions are placed adjacent to the symbols.
Activity durations can also be placed on the chart. Variants on the theme
in Figure 3.38 occur, for example, when work is divided or reprocessed, or
alternative means are possible to perform an activity; such variants lead to
branching, loops and multiple parallel paths respectively.
A flow process chart extends the outline process chart to include TRANS-
PORT, DELAY and STORAGE activities. The chart may be drawn from
the point of view of the worker or the equipment or material used by
the worker. It is similar in appearance to the outline process chart. Travel
distances can be included to scale or by annotation.
Figures 3.39 shows an example flow process chart. Figure 3.40 shows an
alternative flow process chart.
References
Currie, R. M. (1959), Work Study, Sir Isaac Pitman and Sons Ltd, London.
ILO (International Labour Office) (1969), An Introduction to Work Study,
International Labour Office, Geneva.
Description
"
"
Description
"
"
"
"
"
"
Main work
Peripheral/
subsidiary work
Figure 3.38 Example outline process chart.
96 Conventional treatment from a systems perspective
Load
Check
Manoeuvre
Travel Wait Unload
Check
Stockpile
Wait

Figure 3.39 Example flow process chart to emphasise distances travelled.
Wait
Load
Check
Manoeuvre
Travel
Wait
Unload
Check
Stockpile
2 2 1 2 2
Remarks
Total
Activity

Figure 3.40 Example flow process chart.
Production, production rate
‘Production’ is used in this book in a general sense to indicate some measure
of work done. Production rate is this work done per unit of time. The unit
of measurement of production can be various, for example m
2
, m or items.
Production rate may refer to a resource’s constituent behaviour, or when
combined with resource numbers to that achieved in an activity or project.
Program
A program or schedule conveys when work is to be done (and hence it fol-
lows, when resources and money are needed). It gives the time relationship
of activities, and the sequence of activities needed to achieve the desired
project end point. A program contains information at both the activity and
Planning tools and terminology A–Z 97
project levels. However, enlargements upon this meaning will be found in
the literature.
Amongst planners, the term ‘program’ or ‘schedule’ may be applied to
any or all of the following:

(Connected) bar chart

Time-scaled network diagram

Cumulative production plot

Activity date/time listings/timetable

Marked-up drawings.
They typically derive from a network analysis.
A program is presented for the purpose intended. The user has a need for
the quick retrieval of relevant information.
The use of the term ‘program’ here is not to be confused with a computer
program; computer programs (commercial software packages) may however
be used to generate project programs. The use is also not to be confused
with a ‘program’ referring to a collection of projects; that is a project is a
subsystem of a program.
Programs show referenced (as-planned) baselines against which actual
progress can be compared. Programs are continually updated, throughout
a project, to reflect the latest information available. Sometimes bad practice
is seen in that the program becomes an end in itself, and the replanning
loop is not closed.
Project
A good definition of a project is hard to find. (Carmichael, 2004). The
definition, ‘Any undertaking, or set of activities, with starting and ending
points, and with defined objectives and constraints, and resource consump-
tion’ is useful, though perhaps unsatisfactory in a number of ways. Finding
a satisfying definition of a project is difficult.
Perhaps a more suitable way of thinking about a project would be in
terms of attributes that are characteristic of projects, including that they
are:

Unique (one-off, specific discrete undertaking with a unique environ-
ment and unique constraints).

Finite (definable start and end points).
Subprojects. A project is a system and a system can be decomposed to
interacting subsystems which are themselves systems. Likewise a project can
be decomposed to interacting subprojects (Figure 3.41). Each subproject,
which is itself a project, looks after a part of the total project.
Refer Carmichael (2004).
98 Conventional treatment from a systems perspective
Project
Subproject
Sub-subproject
Figure 3.41 Decomposition of a project.
Project management
Agood definition of project management is hard to find (Carmichael, 2004).
Many published definitions of project management are partly circular
describing management as management. Most definitions are usually in
terms of what a project manager does. They often come from studies made
of actual projects and observations carried out on the project managers.
There are three favoured ways of describing what a project manager does,
but other ways do exist:
(i) A ‘classical’ management approach. Management is commonly
described in general management texts in terms of:

‘Planning’

Organising

Staffing

Directing

‘Controlling’

Coordinating.
(ii) Amanagement function approach. An alternative viewof project man-
agement is to regard it as a collection or integration of subfunctions:

Scope management

Quality management

‘Time’ management
Planning tools and terminology A–Z 99

Cost management

Risk management

Contract/procurement management

Human resources management

Communication management.
(iii) Chronological approach. A project may be seen as going through a
number of phases. There is no consensus as to the naming of phases
or to the number of phases. In terms of management, many people feel
comfortable with a chronological way of describing a project and the
activities involved.
Some people find such breakdowns useful. Others find the first two
breakdowns so unsystematic as to be of no use. ‘Controlling’ in (i) above
is in the loose lay meaning sense of the word.
Replanning
Replanning gives revised values of the control variables (method, resources
and resource production rates) for the remainder of the project. The term
‘replanning’ is preferred to ‘project control’ in this book, because lay sloppy
usage of the latter has a tendency to mislead, and is inconsistent with the
optimum control systems usage preferred in this book.
Reporting
Reporting refers to communicating the results of project monitoring. That
is, project output or performance is reported, perhaps in a summarised
form. Reporting can take many forms. The form of reporting is chosen to
match the intended recipient.
Reserve
Also called contingency. Two types of reserve are commonly used – cost
reserve and ‘time’ reserve. A reserve is an extra amount (in money or time
period) added to an initial cost or duration estimate respectively to take
care of possible unknowns.
Resource
Resources refer to people and equipment/plant undertaking the production.
They are distinguished frommoney which is a common unit of measurement
of resource usage (Figures 3.42 and 3.43). They are also distinguished from
materials which refer to that which is incorporated into whatever is being
produced; materials include parts and components.
100 Conventional treatment from a systems perspective
Resources
People
Equipment
Common unit of
measurement
Money
Figure 3.42 Preferred reference to resources.
Resource usage
Money
Figure 3.43 Some possible relationships between resource usage and money.
In this book, the terminology ‘resources’ may be used to refer to resources
generally, to resource numbers (quantity) or resource types. The term
‘resource types’ is used where a distinction between the various resources is
considered necessary. The term ‘resource number (quantity)’ is used where
a distinction within a resource type is considered necessary.
Some people refer (inappropriately in planning) to time, overheads, skills
etc. as resources.
Resource balancing
‘Resource balancing’ is a term applied to resourcing work, and typically
resourcing sequential/repetitive work. Two extremes in handling resources
are possible:
1. Resource numbers (quantity) are chosen to give a desired activity pro-
duction rate. Project duration is considered a higher priority than project
cost. Where multiple activities occur (sequentially or otherwise), it may
be desirable to have the same activity production rates for each activity;
under suchcircumstances, plots for eachactivity ona cumulative produc-
tionplot areparallel, withor without abuffer betweenthem(Figure3.44).
Planning tools and terminology A–Z 101
2. The production rate for each activity is determined by the available or
usual resourcing numbers (quantity). Project cost is considered a higher
priority than project duration. Where multiple activities occur (sequen-
tially or otherwise), plots for each activity on a cumulative production
plot are not parallel; the start or finish times of activities are adjusted
in order to avoid interference between activities, and the project comple-
tion date will relate to the lowest activity production rate (Figure 3.45).
Excavate trench
Lay pipe
20 days
20 days
Excavate trench
Lay pipe
0 3 20 23
Days
S/S lead time period
3 days
C
h
a
i
n
a
g
e
Figure 3.44 Example activities advancing at the same rate, and S/S relationship.
Excavate trench
Lay pipe
20 days
Excavate trench
Lay pipe
0 20
Days
12 days
8
F/F lead time period
0 days
C
h
a
i
n
a
g
e
Figure 3.45 Example activities advancing at different rates, and F/F relationship.
102 Conventional treatment from a systems perspective
Constituent level and element level controls are adjusted to give desired
activity and project level performance.
Collectively, this thinking may be called the line of balance technique. The
first method of handling resources mentioned above might be termed ‘par-
allel scheduling’. The second method might be termed ‘resource scheduling’.
Many projects have the same sequential work repeating throughout the
project lifetime. An example is the construction of multistorey buildings
where the same trades (for example involving the sequence of plumber, elec-
trician, ceiling fixer, and painter) are repeated floor by floor. Other exam-
ples include pipeline and railway construction and estate housing projects.
Resource planning
Resource planning refers to that involved in deciding the usage of resources
on a project.
The resource planning problem is a subproblem of the total planning
problem, where emphasis is placed on the different level controls that can
influence resource usage, and all other controls are held fixed.
Resource plot
Also called a resource profile or resource histogram. A resource plot gives
resource requirements versus time, over the duration of the project.
Days
Days
L
a
b
o
u
r
e
r
s
C
a
r
p
e
n
t
e
r
s
Figure 3.46 Example resource plots.
Planning tools and terminology A–Z 103
The resource plot is developed for each resource type by summing up the
daily (or whatever time unit is applicable) resource requirements for each
activity, as scheduled in a bar chart. This gives the total resource require-
ments on any day. Plotting these values gives diagrams like Figure 3.46.
The diagrams in Figure 3.46 are project level representations. Both
resource smoothing and resource constrained scheduling practices use the
resource plot as their starting point.
Resource smoothing; resource constrained
scheduling
Problems associated with resource usage may be categorised as:

Resource smoothing (also called levelling)

Resource constrained scheduling.
The resource smoothing problem and the resource constrained scheduling
problem are subproblems of the total planning problem, where emphasis is
placed on the different level controls that can influence resource usage, and
all other controls are held fixed.
Generally, when activities are conducted simultaneously, then this leads
to simultaneous demand for resources, producing peak resource demands
at certain stages of the project. Peak demands of resources, particularly over
short time periods, may be undesirable.
Materials problems may be similar to those for resources, and may be
handled similarly.
Resource smoothing involves evening out the resource requirements over
the project or over stages of the project, by reducing peak demands for
resource numbers (quantity) and creating a requirement for resources at
other non-peak time periods. Smoothed resource numbers are desirable,
for example, if people are the resource, because it can provide continuity
in the workforce, eliminate undesirable hiring and firing and attendant
industrial and people problems. Similarly it is not desirable to have plant
and equipment being used intermittently (Figure 3.47a).
Smoothing the requirements for plant leads to higher utilisation and hence
more efficient usage of these investments. Peaks in material requirements
are to be avoided if only to avoid the possibility of shortfalls in supply
during these peak time periods.
Resource constrained scheduling involves scheduling activities such that
their resource requirements never exceed available numbers (quantity). For
example, if plant or equipment is the resource, then large numbers of equip-
ment may not be available. Some resource types can be expensive and
limited in number (quantity). Skilled labour is often difficult to obtain and
expensive to fire. That is, typical constraints will relate to upper limits on
104 Conventional treatment from a systems perspective
Time
Resource
Time
Resource
Upper limit
(a)
(b)
Reschedule to remove peaks
Reschedule to remove
constraint violation
Figure 3.47 Schematic of what happens in (a) resource smoothing and (b) resource
constrained scheduling.
Resources are taken into
consideration when
establishing the nature and
order of work. No further
treatment of resources is
necessary.
Resources are ignored (or
resources are assumed to be
available) when establishing
the nature and order of work.
It is then necessary to carry
out resource manipulation
(resource smoothing or
resource constrained
scheduling).
Figure 3.48 Two broad approaches to dealing with resources.
people with certain skills or upper limits on numbers of pieces of equip-
ment. A schedule is clearly not feasible if it requires resource numbers in
excess of those available (Figure 3.47b).
Resource smoothing and resource constrained scheduling may or may
not be significant problems depending on the degree to which resources
are considered in the initial thinking. When planning is carried out in an
iterative analysis fashion, two broad approaches may be observed when
dealing with resources (Figure 3.48).
Planning tools and terminology A–Z 105
Most people seem to do something in between these two extremes. That
is, they consider resources to a certain extent (maybe in their heads) up
front in planning, but also possibly fine tune later on. The chosen work
method itself can act like a constraint on the way resources are handled.
The solutions to both the resource smoothing and resource constrained
scheduling problems are similar, if not always clear cut. Effectively, if the
resource plot is thought of as hills and valleys, an earthmoving exercise
removes the hills and puts the earth into the valleys. In more fundamental
terms, the project controls (method, resources and resource production rates)
are selected to give the desired resource usage over the project duration.
That is, the solution steps to resource smoothing and resource constrained
scheduling problems are no different to any project planning problem.
The solution could be expected to be more complicated when dealing
with multiple resource types simultaneously.
Controls
A common first approach attempting to solve resource-related problems
(in an iterative analysis sense) is to use activity level controls. If activity
level controls don’t work or don’t work completely, then constituent level,
element level and project level approaches may be tried. Of course, controls
can be tried in any order the planner wishes, rather than following any
suggested order. Some example controls follow:
Solution (controls) at the constituent level. Change resource productions
through, for example, incentive schemes, bonuses, carrots and sticks. These
feed into higher level solutions.
Solution (controls) at the element level. Change resource numbers (quan-
tity). Shift work or overtime might be tried. These feed into higher level
solutions.
Solution (controls) at the activity level. A control which is often tried
first is:

Make use of any activity float. The start dates of non-critical activities
are moved. Note that shifting the start dates of critical activities will
extend the project completion date.
If this doesn’t work or doesn’t work completely, then other activity level
approaches may be tried:

Try compressing the durations of certain activities. Compressing the
durations of activities involves shortening the activity durations, but
this usually comes with a penalty of increased cost and perhaps an
additional and more concentrated resource requirement.
106 Conventional treatment from a systems perspective

Try lengthening activities, though this may come with a cost penalty,
sometimes called uneconomical drag out.

Try splitting activities that can be split. Splitting implies starting an
activity, stopping the activity, starting the activity again after a break
and so on.
Solution (controls) at the project level. Example project level controls are:

Extend the project duration. All activities then become non-critical.

Examine alternative methods of work. This will lead to alternative
network logic. Examples include outsourcing or contracting out some
work instead of doing the work in-house, using machinery instead of
labour, and prefabrication rather than in situ work.

Try overlapping relationships between activities. In place of more usual
finish-to-start (F/S) relationships, start-to-start (S/S), finish-to-finish
(F/F), or start-to-finish (S/F) relationships might be possible.
Other activity level and project level approaches are only limited by the
planner’s imagination. Knowledge of the project and the industry will help
develop alternative approaches.
Whether any of these approaches are possible will be determined by the
project and the nature of the project work.
Algorithms
Many of the better computer packages that perform network analysis
come with resource smoothing and resource constrained scheduling options.
The packages generally have inbuilt heuristic algorithms for solving these
resource problems. The heuristic algorithms try something (such as the solu-
tion approaches listed above) (usually in increments of a day, or other time
unit), and if it produces an improvement then more of the same is tried; if
no improvement is obtained, then something else is tried (Figure 3.49).
Within the solution approaches listed above, there is still a need for the
algorithm to prioritise what is tried.
Try something.
If an improvement occurs, try more of the same.
If no improvement occurs, try something else.
Figure 3.49 Heuristic logic of resource handling algorithms.
Planning tools and terminology A–Z 107
If resource usage and activities are to be manipulated, there is no definitive
rule as to which activities should be manipulated and the order in which
they should be manipulated. Activities may be given rankings or priorities
at the discretion of the planner (or computer package) according to any or
all of the following as well as others:

Earliest start time EST (lowest or first occurring).

Earliest finish time EFT (lowest or first occurring).

Latest start time LST (lowest or first occurring).

Latest finish time LFT (lowest or first occurring).

Duration (shortest or longest).

Free float and total float (least or most).

Resource requirement (largest).

Activity number (lowest).
The allocation of resource numbers is carried out according to the selected
priorities. Where activities have equal priority then appeal is made to the
next lower priority to distinguish between the activities.
The above suggestions are possibilities for establishing a heuristic algo-
rithm for solving the resource smoothing and resource constrained schedul-
ing problems. Other suggestions may be found in the technical literature.
Such approaches will generally mean that a satisfactory solution may be
found but one which is not necessarily optimal. Computer solutions should
also be checked for practicality, and so some form of solution that involves
interaction between the planner and the computer package will perhaps lead
to the better overall solutions. All heuristic algorithms require the planner
to attach priorities and a ranking to activities with different attributes. Each
computer package could be expected to give a different solution.
Small projects can often have their resources rescheduled by inspection.
Measures
How does a computer package, that performs resource smoothing and
resource constrained scheduling, know that one resource plot is better than
another? By human eye, one plot can be seen to be better than another,
but a computer package cannot ‘eye’ the plots, rather some quantitative
measure is needed for comparison purposes.
Rescheduling, whether through such algorithms mentioned above or by
hand, leads to different resource plots. Different resource plots may be
compared quantitatively through the following measures:

Sum of squares (moment)

Number of hirings and firings

Peak requirement

Resource utilisation.
108 Conventional treatment from a systems perspective
A resource plot with a low sum of squares, a low number of hirings
and firings, a low peak requirement and a high resource utilisation is
preferred.
For a resource requirement on day l of u(li the sum of squares measure
is given by
SOS =
¸
all l
u(li
2
(A sum of squares calculation is analogous to a statistical variance [standard
deviation squared] [second moment] style calculation. Compare the shape
of a histogram or probability mass/density function with a high variance to
one with a low variance. Compare the moment of inertia or second moment
of area of a structural member’s cross section, for example an I-shaped
cross section with that of a rectangular cross section.)
The number of hirings and firings is the sum of the vertical portions of the
resource plot. The peak requirement is the maximum resource requirement.
The resource utilisation is the ratio of the area under the resource plot
to the area under a rectangle drawn with one side as the peak resource
requirement and the other side as the time axis.
Multiple resource types
When multiple resource types are present, weighted measures can be used.
The weightings are chosen subjectively to reflect the relative importance of
the resource types.
The weights take account of the relative magnitudes of the resource types.
For example, the average requirements for two resource types might be
2 and 20 respectively; the weightings would be expected to reflect this
difference in magnitudes.
A weighted sum of squares measure would look like,
Weighted SOS = u
1
SOS
1
+u
2
SOS
2
+. . .
where
u
t
weighting for resource type i, 1 =1, 2, . . .
SOS
t
sum of squares for resource type i, 1 =1, 2, . . .
Similar weighted measures can be devised for the other measures of hir-
ings/firings, resource utilisation and peak requirement.
Planning tools and terminology A–Z 109
The final resource plots depend on the values chosen for the weighting
coefficients. It is only the relative values of the weighting coefficients which
are important, not their absolute values. The final solution is a compromise
solution between what is best for each resource considered singly.
Some authors suggest that the weighting coefficients be chosen so as to
give a minimum cost solution, cost here being defined for each resource type
as a unit resource cost (resource cost/day) plus a hiring/firing cost and the
total cost being a sum over all resource types. However, this cost is problem
dependent and generalisations from one problem to another are difficult
to make. Hence to adopt such an approach for obtaining the weighting
coefficients, a whole range of combinations of weighting coefficients has to
be tried and the resource types smoothed and the cost evaluated for each
case. The combination or combinations of weighting coefficients yielding
the least cost solution is/are then used. Restricting the coefficients to integer
values decreases the number of combinations that have to be searched.
However, the number of combinations is still very large.
General
In general, the measures (for example, sum of squares) resulting from mul-
tiple resource types may be treated in several ways:

The measures for all the resource types are combined into a single
measure. The above use of weightings is an example of this.

One resource measure is regarded as more important than the others,
and the less important measures are converted to constraints.

A solution is obtained for each resource measure separately, and these
are then traded off against each other.
Such approaches are no more than is done in multicriteria (multiobjective)
decision making or optimisation (Carmichael, 2004).
Responsibility matrix
Alternatively called responsibility chart. For each activity, there are atten-
dant resource requirements. People resource and organisational (including
coordination) requirements, responsibilities and authorities can be repre-
sented in a responsibility matrix form which may have rows corresponding
to the activities, columns corresponding to resource types and entries cor-
responding to responsibilities, or similar.
The matrices are a way of informing project participants of their roles
and duties.
110 Conventional treatment from a systems perspective
Risk
The exposure to the chance of occurrences of events adversely or favourably
affecting the project/business/. . . as a consequence of uncertainty.
Risk = f (Uncertainty of event, Potential loss/gain from event).
Risk management
The risk management process generally follows Figure 3.50.
Consequence,
outcome
(magnitude,
scale)
Risks
(high/low,
yes/no, . . .)
Uncertainty
Problem
definition
Criteria
Analysis
leads to
Evaluation
leads to
Objectives
Event,
source,
factor
Response,
treatment
Monitoring,
review
Feeds back to source,
uncertainty, consequence
Figure 3.50 The risk management process.
Defining the problem
Selecting objectives
Generating ideas, alternatives
Analysing ideas, alternatives
Selecting the best alternative
Establishing the context
Identifying risk events
Analysing risk situation
Responding
Evaluation of consequences
Problem solving process Risk management process
Figure 3.51 Problem-solving process and risk management process compared
(Carmichael, 2004).
Planning tools and terminology A–Z 111
The risk management process is essentially the same as that used in
systematic problem solving, though different terminology might be used
and the number of step classifications might be different. Refer Carmichael
(2004) and Figure 3.51.
S curve
An S curve is a cumulative representation of either a resource (people or
equipment) or money over the project duration. The curve shape is like a
stylised S for large projects (Figure 3.52). For small projects, the S shape may
not be readily apparent. The S curve is an industry-accepted representation
at the project level. The flatter portions correspond to the project starting up
and the project winding down, with the section in between corresponding
to ‘steady state’ progress of the project, where most resource and money
usage occurs.
The basis for developing S curves is a bar chart. From the bar chart the
resource (and money) requirements can be summed (integrated) over all
activities on a day-by-day basis (or other applicable time unit). These daily
figures (for each resource type, called a resource plot, histogram or profile;
for money, called an expenditure plot or diagram) are then accumulated,
going from project start to project end, to give the S curve (Figure 3.53).
(The cumulative plot is obtained from the resource plot or expenditure
plot in the same way a cumulative distribution function is obtained from
a probability mass function or probability density function in probability
theory. Those people familiar with soil sieve analysis will also recognise the
steps of getting to the cumulative plot.)
Project
start
Project
completion
Time
Initial setting up
of project
Winding down
of project
100%
0%
Project
'full steam
ahead'
C
u
m
u
l
a
t
i
v
e

r
e
s
o
u
r
c
e
/
m
o
n
e
y
Figure 3.52 Example S curve.
112 Conventional treatment from a systems perspective
0%
Days 0
0
Project start
Expenditure
$
100%
Cumulative
expenditure
Project end
Figure 3.53 Example expenditure (upper diagram) and cumulative expenditure
(lower diagram) plots.
Because the bar chart is not a unique entity for any project, for example,
some activities have float and can be adjusted in their start times (between
earliest start times EST and latest start times LST), so a number of S curves
are possible for any project (Figure 3.54) for each resource type and money.
For planning purposes, the particular S curve, which leads to the more
agreeable resource and financial commitments, would generally be preferred
and could be expected to lie somewhere between the two activity starting
time extremes. The S curve for money may be referred to as a cumulative
expenditure plot or cumulative cost plot. It is an industry-accepted way
of representing planned expenditure at the project level, as opposed to
an expenditure plot or diagram that is equivalent to a resource plot in
presentation.
Cumulative expenditure plots are repeatable to a reasonable accuracy
between similar projects, and may be used for budgeting purposes on second
and subsequent similar projects.
The S curve also finds use in reporting for replanning purposes.
Planning tools and terminology A–Z 113
EST S curve
LST S curve
Envelope for
possible
S curves
0%
100%
0%
Percentage of project time
100%
P
e
r
c
e
n
t
a
g
e

o
f

r
e
s
o
u
r
c
e
o
r

m
o
n
e
y
Final planned
usage
Figure 3.54 Envelope of S curves based on activity start times.
Schedule
Refer ‘program’.
Scheduling
Establishing the timing and order of work, that is when work is to be
done (and hence it follows, when resources and money are needed), is
usually referred to as scheduling or programming. However, some people
enlarge upon this meaning and loosely use the term interchangeably with
planning.
Scope
Scope is what is involved in undertaking the project, the extent of the
work to be undertaken, what work is contained in the project (perhaps by
defining what work is not included in the project, in effect the boundaries
to the project) that leads to the end-product. Scope is fully described by
listing all the activities.
There is much confusion over the usage of the term ‘scope’. Some writers
wrongly include extra things besides project work. While useful information
to have, it is not scope, and should be clearly distinguished from scope. The
biggest transgression though is the sloppy, interchangeable use of the terms
‘objective’, ‘constraint’ and ‘scope’; few people appreciate the distinction
and why there is a distinction. Commonly the carriage is put before the
114 Conventional treatment from a systems perspective
horse, such that the scope is said to determine the objectives and the con-
straints. This confusion and imprecision stifles the understanding of project
management.
Refer Carmichael (2004).
Stage
A stage corresponds to a subinterval of the project duration. It may have
physical interpretation or not.
State
The state describes internal systembehaviour. For the stripped-back projects
considered in this book, output and state are the same, that is there are
no observability (in the sense of Kalman) issues. State contains multiple
parts (that is, it is a vector quantity of state variables). State is a controlled
variable.
Task
Refer ‘activity’.
Time management
The term ‘project time management’ (that is, the management of anything
involving time on a project, and including program management) is not
favoured in this book. There also exists the term ‘personal time manage-
ment’ relating to how people ‘use their time’. Neither form actually manages
time, but the terminology is widespread. Also, some people loosely refer to
planning and time management synonymously.
In this book, time is viewed as the independent variable, and projects
are regarded as dynamic systems because of this. The term ‘time’ is used
correctly when used in a technical sense, but is used in a lay person’s sense
elsewhere.
Time-chainage chart
Refer ‘cumulative production plot’.
Time-scaled network
A time-scaled network is an activity-on-link network drawn to a horizontal
timescale. There is no vertical scale. The lengths of the activities (links)
are made proportional to the activities’ durations. A time-scaled network
Planning tools and terminology A–Z 115
gives the same information as a connected bar chart, namely when activities
begin, when activities end, how long activities take, the amount of float
in activities, and the dependence between activities. The horizontal scale is
ideally matched to a calendar allowing for stipulated work hours per day,
weekends, public holidays, rostered days off (RDO), number of shifts per
day and so on.
A time-scaled network may be referred to as a program or a schedule.
A time-scaled network follows directly from a network analysis. The
information from a network analysis is readily transferable into a time-
scaled network form.
Figure 3.55 gives an example time-scaled network.
As with networks and bar charts, different drafting styles are encountered.
For example, the vertical lines in Figure 3.55 might be drawn inclined.
Planners use whichever style appeals to them.
In time-scaled networks, activity durations are given by the horizontal
components of lines; the vertical components have no consequence. Float
is usually indicated by lines with texture different to that used for the
scheduled duration of activities.
Only activity-on-link networks can be drawn to a timescale. Activity-on-
node networks do not lend themselves directly to the time-scaled version.
However this does not stop some people from pretending to draw activity-
on-node diagrams to a timescale. Typically such diagrams end up looking
like Figure 3.56.
10 20 30 40 50 60 80 Days 0
Prelim.
design
Det. dsgn,
base cpts
Deliver
base cpts
Assemble
base cpts
Apply
finishings
Manuf.
base cpts
Approval,
base cpts
Approval,
options
Det. dsgn,
options
Assemble
options
Marketing
Sales
Scheduled duration
Float or dummy activity
Figure 3.55 Time-scaled network for an example project. See the associated bar
chart, Figure 3.1 and networks, Figure 3.25(a,b). (Activity names have
been abbreviated because of space limitations – work items are implied.)
116 Conventional treatment from a systems perspective
Activity A
Duration 10
Activity B
Duration 5
Figure 3.56 Pretend time-scaled activity-on-node diagram.
A true time-scaled activity-on-node diagram would have elongated and
contracted nodes (circles, rectangles, . . . ), and would look silly. Instead,
diagrams like Figure 3.56 are drawn, but these are activity-on-link diagrams
disguised to look like activity-on-node diagrams.
So, when a planner uses a connected bar chart or a time-scaled network,
the planner is unwittingly using an activity-on-link diagram representation,
even though that planner might confess to being a staunch advocate of, or
show a biased preference for, activity-on-node diagrams over activity-on-
link diagrams.
Time-scaled networks also find use in reporting for replanning purposes.
Timetable
A timetable is a document showing when events occur. It may be referred
to as a program or a schedule.
Value management/analysis/engineering
Value management (equivalently value analysis, value engineering) is pop-
ular in project management for a diverse range of situations. Originally
conceived as a way of economising on resources or finding alternatives, it
finds application at all phases of a project from concept through to termi-
nation. Central to the approach is a systematic analysis of function and an
examination of alternatives. As such it can be shown to be but a special
case of systems engineering/problem-solving methodology.
Its origins trace back to about the 1950s when it was known as value
analysis or value engineering. Then it was a design review or ‘second
look’ approach to proposed or existing designs. Its area of application has
enlarged over the intervening years such that it now encompasses not only
design reviews but also, for example, feasibility studies, an examination of
project goals, and conflict situations.
Constructability or buildability studies are but special cases of value
management applied to construction or building projects.
Central to value management is the analysis of function from a whole sys-
tem viewpoint and the proposing or generating of alternatives. It identifies
Planning tools and terminology A–Z 117
wastage, duplication and unnecessary expenditure, and can assist in test-
ing assumptions and needs. The whole system or holistic approach avoids
conventional compartmentalised thinking and obtaining locally optimal or
suboptimal solutions at the expense of the desirable globally optimum
solution.
Refer Carmichael (2004).
Variance
A variance is the difference between planned (or as-planned) and actual
(or as-executed, as-built, as-constructed, as-made, . . . ). It is not referring
to any probabilistic measure of variability.
Variance =Planned−Actual
The usual convention is for a variance to be negative if a project is behind
production/schedule or, for example, over budget. A variance is posi-
tive if the project is ahead of production/schedule or, for example, under
budget.
Using earned value, several types of variances can be identified, including
cost variance and production/schedule variance. There, special definitions
of variance are used.
Work breakdown
A project (scope) can be decomposed into subprojects and then into sub-
subprojects and so on down to the activity level. The lower levels represent
finer detail. The process of decomposing a project (scope) might be called
scope delineation (or scope definition) by some writers. The term work
breakdown structure (WBS) is used to refer to the resulting systematic
decomposition (as opposed to an unstructured activity list where there is a
possibility that some activities have not been thought of).
Work breakdown is not unique; various work breakdowns are possible
for any project. A breakdown is chosen that makes life as easy as possible
for the planner. The degree of coarseness or fineness with which a project
is subdivided will depend on the end use of the plan. There is no right or
wrong answer to this, only better or worse subdivisions.
Some people (wrongly) see this breaking down as going from the ‘product
to the processes’, or as a product-oriented tree. However, since a project
is not a ‘product’, what is happening is more akin to generally defined
processes being broken down into more refined processes.
Some writers (wrongly) say that scope delineation involves subdividing
the project deliverables. (Deliverables are the end-product plus project state
final conditions.) But only project work should be involved in the break-
down, and not the end-product.
118 Conventional treatment from a systems perspective
Planners distinguish between events and activities. Each activity has
events associated with it. Planning with either is okay, but care has to be
exercised when the two are mixed. Some writers (wrongly) include events
in work breakdown structures.
Terminology for each of the levels, used by different people, might include
project, subproject, work package, job, subjob, process or activity. The
terminology of task and subtask might also be used at the middle to lower
levels. There is no consensus in the use of terminology to describe the
various levels, though this book’s preference is for the terminology project –
subproject – . . . – activity. ‘Program’ is a term used to refer to the level
above projects and represents a collection of projects; that is a project is
a subsystem of a ‘program’. An alternative usage of the term ‘program’ is
given above.
Work method
Work method is the sequence or logic of operations, with activity and event
interrelationships. It is commonly represented by a network.
Work study
Work study attempts to create in people a questioning state of mind in
the way they view their current and future work practices. This aware
state of mind applies to all phases of a project and on a continuing
basis in an organisation. Work study involves the critical and systematic
analysis of work with an ultimate view to improvement and the elim-
ination of any nonproductive components. Work study is made up of
method study and work measurement. Method study breaks down work
into its components and questions the purpose and need of each com-
ponent; it involves, amongst other things, recording information on the
work, critically examining the facts and the sequence and developing alter-
natives. Work measurement is concerned with the time periods taken to
perform work. Work study is an established technique that has wide
applicability.
Work study may be broadly defined as an examination of the use of
people, equipment and materials in work tasks, and an associated attempt
at improvement and elimination of waste. Besides the financial incentives,
it attempts to create an attitude of mind about the effective use of people,
equipment and materials.
Work study can be shown to be a special case of systems engineering/
problem-solving methodology. Re-engineering is work study reinvented
(Carmichael, 2004).
Planning tools and terminology A–Z 119
Exercises
1. Under what circumstances should incomes and expenditures, as shown
in Figure 3.17, be plotted based on dates of anticipated transfer of
funds and billings, or based on invoiced, committed or paid amounts, or
other?
2. (a) Isolate the parts of the ‘classical’ management function approach
of what are categorised as ‘planning’, organising, staffing, direct-
ing, ‘controlling’ and coordinating that imply involvement with (i)
schedule/production, (ii) money, and (iii) resources. How might
this ‘classical’ management function division help or hinder the
understanding of planning as outlined in this book?
(b) Isolate the parts of the management function approach of scope
management, quality management, ‘time’ management, cost man-
agement, risk management, contract/procurement management,
human resources management and communication management
that imply involvement with (i) schedule/production, (ii) money
and (iii) resources. How might this management function division
help or hinder the understanding of planning?
3. Some planners include in their program an allowance for delays – due to
weather, industrial action or other unforeseen circumstances. These delays
are incorporated into programs as separate activities. The argument behind
this is that the overall program should present as realistic as possible picture
of the duration of the project.
What do you think of such a practice?
Some planners will include delays in all programs except at the very
detailed level; at the lowest level it is believed that by incorporating delays,
focus on target dates disappears. By including delays does it take away the
emphasis on target dates?
Will knowledge that an allowance for delays has been included in the
program make people think that delays are inevitable or expected? Is this
a case of Parkinson’s Law? – Work expands so as to fill the time available
for its completion. Is this a case which supports the notion of self-fulfilling
prophecies?
Will some of the project team assume they have longer to do their work
because a delay has been included in the program for other members of the
project team – for example, will designers take longer knowing that a delay
has been incorporated into the site work program?
Should delays only be incorporated in the broad program, with detailed
programs omitting delays in order to focus attention on target dates?
Should there be two programs – one without anticipated delays (a target
plan, made public) and the other with delays (kept private)?
120 Conventional treatment from a systems perspective
4. Obtain a user’s manual for a computer programthat does network anal-
ysis including resource handling and overlapping relationships. Research
the program input and program reporting facilities.
5. For the following project network in activity-on-link diagram form,
convert it to an activity-on-node diagram form. For convenience, activities
have been listed as letters of the alphabet.
A
C
B
D
E
F
G
Convert the following project network in activity-on-node diagram form to
an activity-on-link diagram form.
A
B C
D
E
F
G
H
6. Why might some planners dogmatically criticise activity-on-link dia-
grams (while stating a preference for activity-on-node diagrams), yet quite
happily use bar charts?
7. An S curve (money) represents cumulative expenditure. Why is this
form of diagram preferred by industry over reporting on individual activity
expenditure, much like a bar chart reports on individual activity durations?
8. In examining some proposed forestry (timber harvesting) operations,
an environmental impact statement is prepared. The main activities in this
are listed in Table E3.1. The order in which the work is planned to be done
is according to the network logic of Figure E3.1.
Planning tools and terminology A–Z 121
Table E3.1 Forestry study activities
Activity code Activity Normal duration
(days)
Object Establish proposal objectives and need 10
Descrip Develop description of proposal 3
Altern Examine alternatives 2
Physic Study physical environment and the effects of
forestry activity
20
Atmos Study atmospheric environment and potential impacts 6
Hydrol Study hydrological environment and impacts of the
proposal
15
Veget Study vegetation and effects of the forestry activity 20
Fauna Study the fauna and effects of the forestry activity 20
Econo Examine economics and land-use effects 15
Cultur Examine cultural environment and effects of the
forestry activity
10
Coord Coordinate project 25
Report Produce report 55
Object Descrip Altern Physic
Atmos
Hydrol
Veget
Fauna
Econo
Cultur Coord
Report
Figure E3.1 Forestry study logic.
Assume each of the ‘Study’ activities may be reduced in duration by 20%
and the remaining activities by 10% (each rounded up to the nearest day).
What is the possible shortest duration?
Chapter 4
The planning process
Introduction
Planning dimensions
Planning starts with broad assumptions which are steadily refined, as more
information comes to hand. Planning typically proceeds from the broad to
the detailed.
Planning is typically done from the general to the particular, from coarse
to fine, in two directions (or two dimensions) (Figure 4.1) at the same time:

In hierarchical levels involving breaking (the work in) projects into
subprojects (for subprojects other than that based on stages), sub-
subprojects, . . . , activities, elements and constituents. Detail emerges
with lower levels. Project work is dissected into manageable portions
through work breakdown (discussed later).

Over time. Detail evolves as the project progresses.
The launching point for the main project planning is at the time of estab-
lishing the initial scope (Figure 4.1), which derives from the objectives and
constraints. For each and all levels and at any time in Figure 4.1, a planning
problem exists.
Aplan near the top left of Figure 4.1 might be called a ‘project preliminary
plan’ or similar name. It contains broad information on work method,
resources, dates and so on. And it may be used as a basis for approval to
do further project work, as might other plans at later stages.
As a synthesis problem
Planning is a synthesis problem. As such, there are multiple solutions
(choices of control) possible. In most cases, planners are only after a sat-
isfactory solution, or a solution that they can live with, and do not put
The planning process 123
Fine
Coarse
Coarse
Fine
Level
Stage
Initial
scope
Final
scope
Detail evolves over time
D
e
t
a
i
l

u
n
f
o
l
d
s

w
i
t
h

p
r
o
j
e
c
t
,
s
u
b
p
r
o
j
e
c
t
,

.

.

.

,

a
c
t
i
v
i
t
y
,
e
l
e
m
e
n
t

a
n
d

c
o
n
s
t
i
t
u
e
n
t
Figure 4.1 Evolution of planning problems.
additional effort into searching for the optimal solution. A planner may
also be under pressure to come up with quick solutions.
Planners are unaware of and do not understand the components of the
synthesis problem, and so they never know where they are relative to the
optimum. They are unable to vocalise or formulate the synthesis problem
components. Such discussion goes to the very heart of current planning
knowledge being in its infancy, and current planning education (read train-
ing) being superficial and low standard.
As an iterative analysis problem
The solution of any planning problemvia an iterative analysis mode involves
the conventional steps in systematic problem solving or systems engineering.
Feedback occurs between steps in trying to refine the problem. (This is
additional to the feedback that can occur between levels.)
Alternatives might be arrived at through the generation techniques used in
a value management/analysis/engineering study, which aims at developing
alternatives that perform a desired function but at a lower cost, or in a work
study, which is concerned with the method and timing of work activities
and their optimisation. Both value management and work study are no
more than versions of systematic problem solving (Carmichael, 2004).
Case example – Small commercial buildings
This case example looks at one contractor’s procedures for plan-
ning the construction of small commercial buildings involving minor
124 Conventional treatment from a systems perspective
earthworks, concrete work, stormwater drainage, steel fabrica-
tion/erection and electrical fitout.
Three phases/types of planning – initial, detailed and ongoing – are
used.
The initial or preliminary planning occurs during preparation of a
tender. A program is developed for submission with the tender and to
assist the estimating department inpricing the project; the programmay
be required to assist the owner to further assess the contractor’s offer.
The project duration may be set by the owner and it is a case
of allocating sufficient duration to each activity so that the overall
project is completed within the nominated duration.
The detailed planning does not occur until the tender has been
successful. When a contract has been awarded, and within the first
couple of weeks, a detailed construction program is developed. It is
commonly a contract requirement that a construction program be
submitted within 14 or 21 or 28 days of contract commencement.
Work on the construction program incorporates a review of the ini-
tial program, specifications and drawings. After possible discussions
with the site foreman, it is determined which activities of the initial
program can be retained and which activities require further breaking
down. The aim is to develop a program with sufficient detail that
can be easily followed and updated. Experience helps in determin-
ing activity durations and overlapping relationships to complete the
project in the nominated time frame. This information is then input
to a commercial project management computer package so that the
critical paths, floats etc. can be determined and so as to have the pro-
gram in a professional form for submission. The program is presented
as a connected bar chart.
Ongoing planning of the project leads to updating the construc-
tion program on a monthly basis and the preparation of fortnightly
programs (programs for the following two weeks). The fortnightly
programs contain considerably more detail than the construction pro-
gram and are usually updated on a weekly basis.
Planning problem components
The main components of any (optimal) synthesis problem (Figure 4.2), to
which planning belongs, are (Carmichael, 1981):

Model

Objective(s)

Constraints.
The planning process 125
Optimal synthesis
(inc. planning)
problem components
Model
Objective(s)
Constraints
Figure 4.2 Optimal synthesis and planning problem components.
Objectives and constraints are developed below. Models are developed
separately (Figure 4.3):

Hierarchical models – Chapter 5, Chapter 9

Multistage models – Chapter 8. The project is broken down into stages,
or equivalently the project duration is discretised

Probabilistic models – Chapter 7.
These problem components are expressed in terms of control, state and out-
put variables. The system notions of control, state and output variables are
central to the development of models and thinking for planning purposes.
Models
Hierarchical models
Multistage models
Probabilistic models
Figure 4.3 Model development in this book.
126 Conventional treatment from a systems perspective
Control
Control is equivalent to a planning decision, action or input. The term
‘control’ is favoured here in place of decision, action and input. The planner
exerts control on the project in order to bring about a desired project
behaviour. The most obvious controls are work method, resource (people
and equipment) selections and resource production rates. Work method,
resource selections and resource production rates may be freely selected by
the planner, subject to any constraints being present. The preferred controls
are those that extremise the objectives.
State, output
The output describes external or observable response, performance or
behaviour. Internal behaviour is usually given in terms of system state. By
the nature of projects (in the stripped-back form treated in this book), the
internal and external behaviour are the same. (The observation equation
of control systems theory [Carmichael, 1981] becomes ‘output = state’.)
There is a one-to-one transformation between output and state. The terms
‘output’ and ‘state’ become interchangeable when referring to projects, and
there is no noise in the observation device. There are no observability (or
controllability) (in the sense of Kalman) issues for projects thought of in a
stripped-back form.
The most obvious states are the resources used so far (cumulative
resources) or the cost so far (cumulative cost) and the production so far
(cumulative production).
Disturbance
In most systems studies, there is something that prevents the hoped-for out-
put being obtained exactly. Commonly this is because of a changing project
environment brought about because of delays, disruptions, adverse climatic
effects and so on. Collectively the corruption is referred to as disturbance
below, though control texts may use the term ‘noise’ (Figure 4.4).
Disturbance
State
(System)
Control/input Output
Figure 4.4 Schematic representation of control, state and disturbance.
The planning process 127
Example
Planning establishes how and what work will be carried out, in what
order and when and with what resources (additionally expressed in a
money unit).
Consider the simpler (subsystem) problem where it has already been
largely established how the work will be done and with what type
and number of resources, and at what resource production rates. As
such, the controls available to the planner are the order or sequence
of the work including when particular work items will be done.
This establishes when resources are used, or, if expressed in a mon-
etary unit, when money is used.
What the planner is trying to influence is the utilisation of the
resources (state variables), and thereby the duration of the work.
There may be constraints on costs, duration of the work and
resource availability.
Ultimately, what the planner might be trying to achieve is a work
sequence of minimum cost or minimum duration or a compromise
between these two objectives.
A suitable system model is the time-scaled critical path network or
connected bar chart, with development to resource plots and cash-
flow diagram.
As the planner adjusts the sequence of work, so resource utilisations,
alternative cash flows and project durations evolve. By a series of
iterations, the planner settles on one particular sequence of work.
Using an optimal control approach, the components to the planning
problem are:
Model – Time-scaled network, connected bar chart, . . .
Constraints – Related to work duration, costs, resource availabil-
ity and resource production rates.
Objective(s) – (Minimum) cost and/or (minimum) duration.
(After Carmichael, 2004)
Simplifications to planning
Ideally, project planning should be carried out as a probabilistic (inverse)
synthesis problem. However, the tools are not currently available for this
as a stand-alone mathematical problem. The full planning problem and
its solution are very complicated, requiring ingredients of creativity and
128 Conventional treatment from a systems perspective
Planning problem
simplifications
Conversion from
synthesis to
iterative analysis
Hierarchical
breakdown of
a project
Conversion from
probabilistic to
deterministic
Figure 4.5 Simplifications to make the planning problem more tractable.
experience. Instead, planning is simplified and made more mechanical in
three main pragmatic ways as shown in Figure 4.5.

Instead of solving the planning problem in a synthesis format, it is
converted into an iterative analysis format. Analysis is much easier than
synthesis, and if a good initial guess is made, then there may be only
one iteration (or even zero iterations) needed to give a sufficiently good
answer. Iterations may modify the method, resource usage and resource
production rates until satisfactory (without necessarily being optimum).
For most planners, the step of generating alternatives is relatively
straightforward without knowing howclose to the optimumthe guessed
solution is.
In some cases, where the planner is feeling his/her way, the analysis
may become a ‘what if’ or sensitivity analysis examining the impact of
alternative controls.

The project (work) is broken down (hierarchically) to lower level sub-
projects, sub-subprojects, and so on to activities, elements and con-
stituents. Subprojects may be chosen such that the interaction between
subprojects is minimal and is commonly neglected, permitting sub-
projects to be planned independently of other subprojects. The same
holds true where multiple projects are combined on an organisation
level. However, interaction at levels lower than subproject is commonly
strong.
Interaction occurs because of the work method (implying an
order/sequencing of work). Iterations (between levels) can largely be
removed by the planner being aware of what is happening on other lev-
els when solving an individual level problem. The levels are uncoupled
as far as possible, for tractability reasons, to make the level problems
less complicated.

The planning problem is assumed deterministic. The analysis is
made deterministic. Estimates for resources, productions, durations,
costs, . . . are made deterministic in nature. Future events (leading to
both upsides and downsides) are anticipated. Where future events have
The planning process 129
a likelihood (uncertainty), this is incorporated into thinking akin to
risk management; there some probabilistic analysis technique, such as
Monte Carlo simulation, may be used, but more usually is handled
deterministically through a ‘what if’ or sensitivity analysis.
Planning is always carried out over the full project time span. Planning stage
by stage, without consideration of upcoming stages, could be expected to
lead to a suboptimal solution.
Replanning
Replanning is carried out according to stages or at discrete points in time,
but this follows from the nature of what replanning is. Replanning is no
different in form to planning. The planning problem (for the remainder of
the project) becomes better defined as the project progresses. Information
assisting the planning problem could be expected to be better the less far
into the future. Replanning is based on the remaining duration of the project
using the current state of the project as initial conditions. A good plan
allows for some flexibility and can be adapted to changing circumstances
that may occur during a project. However, attempts are made to foresee
potential influences in upcoming stages, and controls are chosen with these
influences in mind. According to Machiavelli in The Prince: When trouble
is sensed well in advance it can easily be remedied; if you wait for it to show
itself any medicine will be too late because the disease will have become
incurable.
Whether replanning is necessary or not is established through monitor-
ing/sensing the project output (‘as-executed’) through time, and comparing
this with the current baseline (‘as-planned’). The difference between the ‘as-
planned’ and ‘as-executed’ values (together with an updated view of future
influences) may then be used to indicate what controls need to be modified,
that is, to help in the generation of alternative controls for replanning.
Planning at some stages may take on increased importance for differ-
ent stakeholders, for example for an owner at project outset, and for a
contractor when tendering and while undertaking the work.
Objectives and constraints
General
Objectives and constraints derive from the value system of the project/
end-product owner. Objectives and constraints lead to scope, and other
planning matters. Scope is selected based on stated objectives and con-
straints. Without defined objectives and constraints, the ensuing scope and
planning is ill-defined.
130 Conventional treatment from a systems perspective
The term ‘planning objectives’ can occasionally be found in writings on
planning, but its intent is not in a systematic problem-solving sense, and
hence is not used here.
The objectives and constraints carry through to all planning problems at
all levels (project, activity, element and constituent) and over time. They
may be decomposed to finer detail and stated more specifically for use in
lower levels, and later times.
Objectives
There will be an identified need or want for some product, facility,
asset, service etc. This end-product is achieved through a project.
The materialisation of the end-product can be performed, possibly,
in an infinite number of ways. In systems studies terms, the problem is
an inverse problem, and hence there is no unique solution.
The criteria by which the preferred materialisation of the end-product
is selected are the project objectives. Scope and other planning consid-
erations on projects follow.
In a similar idea, there are possibly an infinite number of versions
of end-products. The criteria by which the preferred end-product is
selected are the end-product objectives. Form, function, finishes etc of
the end-product follow. The selection of the preferred end-product form
is a design problem.
That is, on any project there are two types of objectives

End-product objectives

Project objectives.
See Figure 4.6.
Commonly, project objectives say something about project cost,
project [duration] and deviation from specification, but other objectives
are possible. And these may apply throughout the project (for example
minimum deviation from specification), or at the final (terminal) point
of the project (for example minimum [duration which is related to the
project] completion time). That is, general project objectives will con-
tain a component over the time domain of the project and a component
at the final (terminal) point.
Objectives
End-product
objectives
Project
objectives
Figure 4.6 End-product and project objectives.
The planning process 131
Example issues included in objectives
End-product objectives and project objectives (and constraints)
may relate, for example, to:

Money (end-product – sales, benefit:cost ratio (BCR), net
present value (NPVi, . . . , project – cost, budget, . . . )

Duration (end-product – lifetime, . . . ; project –
duration, . . . )

Resource usage

Quality issues

Community acceptance

Environmental (natural) effects

Safety

Risks

Public impact

Extreme event impact (floods, cyclones, . . . )

Social impacts

Geotechnical considerations.
These are expressed in terms of end-product matters or project
matters, as the case may be.
Confusion can arise when people start expressing project objectives in
terms of the end-product requirements or end-product objectives.
For any situation there may only be a single objective or there may be
many objectives (multi-objective). Where there are multiple objectives,
they may be non-commensurate and conflicting. One or more of the
objectives may be more dominant than the others.
The solution of such multi-objective problems generally involves some
subjectivity while the solution of the single objective problemdoes not.
Constraints
All projects have genuine constraints such as funding, environmental (nat-
ural) and political. Constraints limit the range of controls possible; they
restrict the options that are possible.
As with objectives, many people confuse a project constraint with an
end-product constraint. End-product constraints may influence project con-
straints (Carmichael, 2004).
The higher levels and early project stages constrain the choices in the
lower levels and later project stages respectively. Detailed planning is estab-
lished within the parameters of previous broader planning.
132 Conventional treatment from a systems perspective
Constraints can be converted to objectives (and vice versa).
Example objectives (project, activity, element and/or con-
stituent levels)

(Minimum) impact of activities on the natural environment

(Minimum) project duration

(Maximum) utilisation of resources

(Maximum) return on money invested (in project and end-
product) – related to usage of available money.
Example constraints (project, activity, element and/or con-
stituent levels)

Availability of resources, of the required type and in the required
quantities when needed; a smoothed requirement for resources

Availability of money, in the required quantities when needed

A certain specified amount of interaction with the community

Complete the work for less than (money). Cost. Budget. Spending
limitations. Profitability. Cash flow

Hazards and local factors; safety issues

Statutory, legal requirements; environmental considerations

Allowable sound volumes

Site space restrictions; resource movement, storage, access, work
flow

A lower limit on the standard of work; quality assurance

Designated milestones. Deadlines. Project completion time (may
be imposed or as-planned)

Production/schedule delays due to, for example, fundingapprovals,
regulatoryapprovals, site access limitations, manpower availability
problems and weather problems

Events that are required to happen at the right time

Technical knowledge, capabilities

Transport, delivery issues.
It is frequently the case that a duration constraint or a cost constraint will
be binding in the planning solution. That is, the preferred planning solution
will lie on either the duration constraint or the cost constraint.
The planning process 133
Care has to be exercised that any imposed constraint, especially an unre-
alistic cost constraint, does not prevent the free choice of sensible controls
(method, resources and resource production rates). A certain desired stan-
dard/quality in performance may leave little flexibility in the choice of
resources. In the extreme case, a constraint imposed without thought may
be such that there are no feasible controls. Hence if the project proceeds
in whatever form, the constraint will be violated; this may represent, for
example, a financial loss or a completion time overrun, or the scope may
have to change – ‘you get what you pay for’.
Case example – Fruit processor upgrade
This case example discusses the practices used within a company to
assign resources to a project involving an upgrade of an owner’s exist-
ing fruit processing. The upgrade involved civil, structural, mechanical
and electrical work, and the completion date was fixed by the need
to have the work completed and the fruit processor running when the
first fruit was ready for picking.
The company doing the project work was structured in three sec-
tions along discipline lines – civil/structural, process/mechanical and
electrical/instrumentation.
Once the project manager had determined the resources required
on the project, section heads assigned these resources to the project.
Following scope delineation, a program was developed to show
how the work had to be done in order to meet the end date. The
construction program was developed backwards based on the defined
end date and the delivery dates for some long lead time period items
that had been previously ordered.
The time period before the construction started then became the
time period available for design and documentation, after allowing
a sufficient time period for tendering. The other factor driving the
design phase program was that the project manager was going to site
for the construction phase and, in order to most efficiently run the
design phase, all the documentation had to be completed before the
project manager went to site. This resulted in a six-week time period
in which to design and document the job.
At this stage, based on the scope (of work) for their sections, the
section heads advised the number of hours they required to complete
their work. They were given an indication of the program, and were
asked to advise who they would assign to the job.
The hours given by the sections formed the basis of the fee proposal
presented to the owner, and the owner instructed the company to
proceed and commence design.
134 Conventional treatment from a systems perspective
At this stage the project manager felt that all necessary resources
were in place because all section heads had the hours they had asked
for, and the project was proceeding on schedule.
As the design phase proceeded, it was found that the people who
had been identified to work on the project were being diverted from
this project onto other projects, and the design began to fall behind
schedule.
Discussions with the section heads brought no solution and it was
only after the project manager complained to one of the company
directors that pressure was brought to bear on the section heads and
the project was assigned the resources it needed.
At this point the design was about two weeks behind schedule. For-
tunately, it was possible to get the critical documentation done during
the period of reduced resource availability so that the construction
schedule did not suffer greatly.
Off-setting this partly was a delay in receiving some information
from the owner, which also delayed the documentation for a com-
ponent of the work, and so the lack of resources available for this
particular work package was not a problem.
The project encountered difficulties because it was only one of
a large number of projects which were then happening, and which
required resources from a finite pool.
Scheduling the use of resources could be difficult at the best of times,
because owner approval for projects was often delayed, sometimes
for weeks, and the company had to try and juggle the resources it had
among whatever projects were going at any one time, and hope that
the projects lying in the wings did not all happen at once.
In this case a number of other projects, that had had owner approval
delayed, all got the go ahead just as this project was starting. In
addition, a number of new projects ‘came in the door’, and hence
there was a large demand for resources that the company did have.
The section heads did not carry out any formal resource smoothing
across all the projects requiring resources. Resources seemed to be
assigned on a perceived priority and on a ‘whoever screamed the
loudest’ basis.
Scope
Scope is what is involved in undertaking the project, the extent of the
work to be undertaken, what work is contained in the project (perhaps
by defining what work is not included in the project, in effect the
boundaries to the project) that leads to the end-product. Scope is fully
described by listing all the activities.
The planning process 135
There is much confusion over the usage of the term ‘scope’. There
is much loose usage of the term ‘scope’. The biggest transgression is
the sloppy, interchangeable use of the terms ‘objective’, ‘constraint’
and ‘scope’; few people appreciate the distinction and why there is
a distinction.
Commonly the carriage is put before the horse, such that the scope
is said to determine the objectives and the constraints.
Getting to the scope is the process of working through the inverse
problem associated with the ‘means to the end-product’. Objectives
and constraints are delineated, alternative work methods are proposed
and evaluated against these objectives and constraints, and a best work
method selected. This defines the extent of the work or scope.
As a project progresses, the objectives may change but more likely
the constraints will change (for example, the work completed so far
constrains what can be done next, assumptions on delivery time periods
were inaccurate, . . . ). This leads to scope changes.
The extent of the project work follows from a knowledge of the
project objectives and constraints. Work established without reference
to objectives or constraints could be expected to be non-optimal with
respect to the objectives, and also possibly violate the constraints. Some
people are able to rationalise this as being satisfactory, and some even
rationalise it on the basis that they are being ‘practical’. In reality,
though, such people are unaware of how the scope follows from a
knowledge of the project objectives and constraints.
Attention is paid to scope in order to indirectly [contain] deviations in
cost, [schedule] and quality fromthat planned. Scope changes on projects
are common and may be referred to as variations or extras. Some projects
have a tendency for multiple changes leading towhat is sometimes termed
scope creep. Scope changes may come about through changed thinking,
errors, changed values and circumstances, unforseen issues and so on.
People confuse a change in the end-product with a change in scope;
the first will commonly lead to the second, but they are not the same.
People also wrongly talk of control when they speak of ‘change con-
trol’; invariably what is involved is a monitoring and recording of
changes and perhaps some restrictions on changes. Changes then lead
to replanning of the remaining part of the project.
Scope becomes better delineated as the project progresses. For many
projects it is not possible to knowthe exact scope at the start of a project.
(Carmichael, 2004)
From objectives and constraints to scope
Scope derives from the project objectives and constraints. Figure 4.7 shows
the steps involved. (Planning concerns itself with the ‘means to end-product’
136 Conventional treatment from a systems perspective
End-product
Identify need
for end-product
Creates need
for means
Establish objectives
and constraints
Generate alternatives Generate alternatives
Feasibility/select
preferred end-product
Feasibility/select
preferred means
Broad scope
Establish objectives
and constraints,
deriving from:
(i) End-product
objectives and
constraints
Means to end-product
End-product concept
Design
development
Plan
development
(ii) Elsewhere
Figure 4.7 Broad design and planning steps.
problem. The parallel problem of design concerns itself with the ‘end-
product’ problem.)
The means to the end-product is based on knowledge and experience
relevant to that project’s industry, and an understanding of that indus-
try’s projects. Ideally, all available ideas are evaluated. Viable alterna-
tives/options are assessed including the carrying out of cost–duration
trade-offs. Peculiarities of the project are considered.
Scope and planning
Scope establishment is part of the initial broad planning problem. In
Figure 4.1, this is referring to the top left corner.
The planning process 137
Extras feeding into and out of planning
Into
For planning to take place, supporting information is needed to feed into the
basic problem-solving steps. Supporting information includes (Figure 4.8):

Assembled and identified project data, for example, that obtained
through a site inspection. Some familiarisation of the project environ-
ment and the nature of the work commonly takes place.

Estimates of activity durations, associated resources (people and equip-
ment), resource production rates and money to carry out these activi-
ties – availability of resources and money in time (when) and quantity
(what). Estimates can be refined as the project progresses, and more
information comes to hand; this in turn can lead to more accurate and
more detailed plans.

The behaviour of people (Chapter 7) including people’s motivation,
resistance to change and work ethos.

Potential difficulties are anticipated. Forecasts are made and assump-
tions behind forecasts are tested.
Communication
The outcome of planning, together with the underlying planning assump-
tions, are communicated to the project stakeholders. Various forms of com-
munication are possible, usually as documents (but can include physical
and computer models) and verbal communication. The documents provide
a basic reference and briefing for those who will ultimately be involved
in the project execution. The documents allow project outsiders (including
senior management) to review the plan, provide guidance and assist and,
thereafter, support the plan throughout project execution.
If, following review, the plan is found to be satisfactory, the plan is
implemented.
Planning
Project data,
Estimates,
Availabilities,
People behaviour,
. . .
Programs,
Communication
of the plan
and underlying
assumptions,
. . .
Figure 4.8 Extras feeding into and out of planning.
138 Conventional treatment from a systems perspective
Table 4.1 Commonly used graphical forms of communication
Information with respect to Communication means
Schedule Bar chart, time-scaled network
Resources (people and equipment) Resource plot, list; cumulative resource
plot (S curve)
Money Cumulative money plot (S curve), cash
flow diagram, list
Production, work done; activity/project
production rates; material usage
Cumulative production plot
Delays Cumulative delay plot
There are a number of useful ways to assist the planner organise and
present planning information. These take the form of charts, plots, activity
date/time listings and so on. They might be regarded as the end-result of
the planning process to date. Commonly used graphical forms of commu-
nication are indicated in Table 4.1 (refer Chapters 3 and 6 for descriptions
of these).
As communication tools, there is a requirement for clarity, explicitness,
being uncluttered, without superfluous information, and ease of under-
standing (intelligible) for all the end-users. Covering reports can be used
to state assumptions, calculations and so on, and to keep the diagrams
uncluttered. Colour can be used to improve interpretation.
According to Murphy’s Law (extended): The most important piece of
information in any plan or document stands the greatest chance of being
left out.
Programs
General
Programs are a means of communicating the outcomes of planning. Ideally,
some comment on the underlying planning assumptions accompanies the
program, in order that end-users are well-informed, and can see the bigger
picture. The forms programs take reflect what is trying to be communicated.
Programs may be any of the following:

(Connected) bar chart

Time-scaled network diagram

Cumulative production plot

Activity date/time listings/timetable

Marked-up drawings.
The planning process 139
They ideally derive from a critical path analysis. They may commonly be
presented:

Over different project levels

Over different time horizons.
Associated with such programs, there may be a need to also communicate
resource and money requirements in the form of:

Resource plots; S (resource) curves

Budget, cash flow; S (money) curve.
and also possibly delay information.
End-users are acquainted with their individual responsibilities, possibly
through responsibility matrices.
See Chapter 3 for the descriptions of these representations and termino-
logy.
The program is the visible part of the planning process, used to commu-
nicate the results of planning. Programs are guides to assist the project work
being carried out efficiently. They show the steps involved in the project
work. Project participants can use the programs to see where they fit in the
project. Durations and start dates might be arbitrarily imposed or decided
in cooperation with all stakeholders, including contractors and equipment
suppliers.
Programs give referenced (as-planned) baselines against which actual
progress can be compared. Programs are continually updated, throughout
a project, to reflect the latest information available. Sometimes bad practice
is seen whereby a program becomes an end in itself, and is not used to assist
replanning and project management.
For large projects, the number of activities may become unwieldy to han-
dle together, and it may become necessary to create a number of convenient
subprojects, or to group activities.
It is possibly too obvious to say that programs should be appropriate
to the relevant end-user, and that they should be clear, user-friendly, at
the appropriate degree of detail, and uncluttered in order to communicate
the agreed plan. Some planners use the K.I.S.S. (keep it simple stupid)
principle in order to improve understanding. Complicated programs are not
recommended.
The maintenance of the program may be the responsibility of the contrac-
tor. Subcontractors may have a say in where they fit in the program, or may
be advised when they are required to appear. Definiteness on these appear-
ance dates assists subcontractors to manage their overall affairs. Vagueness
can lead to a falling out between contractor and subcontractors as each
may have a different view on what is expected of the subcontractors.
140 Conventional treatment from a systems perspective
Programs over levels
On major projects, the communication of planning outcomes in forms
such as programs could be expected to occur on several levels. The levels
roughly align with organisational levels. Senior management may only be
interested in broad issues. Middle management would show interest in more
detail. Individual production groups would work with still more detailed
information. There is no consensus as to the number of levels used or what
each level is called. Figure 4.9 shows three levels, but less or more than this
may be adopted depending on the situation. The higher the level, the lower
the degree of detail. With reference to Figure 4.9:

Upper level For senior management and the owner, a program is
developed showing key work parcels and sequencing, milestones, and
possibly contingencies (Figure 4.10).

Middle level Middle management requires an indication of major
work parcels and major events.

Lower level Production management or workface management works
off a program suitable for individual production groups.
Owners and senior management tend to be more concerned with overall
results and not in the detail. In contrast, a foreman, say, needs detailed
information over short periods to guide the workface employees. Interme-
diate detail may be required for contractual purposes when tendering, and
may show major items and completion times.
Senior management,
owner
Middle management
(Engineering, construction,
commissioning, . . .)
Production management,
workface management
Specific work activities,
associated events
Upper
level
Middle
level
Lower
level
Key work parcels,
milestones
Major work parcels,
major events
Figure 4.9 Example hierarchy of programs.
The planning process 141
J F M A M J J A S O N D J F M
Subproject Responsibility
This year Next year
Venue 1 Venue
Committee/
Secretariat
2 Printing and
Publications
Venue 4 Audio Visual
Equipment
Secretariat 5 Registration
Committee/
Secretariat
6 Speakers/VIPs
Secretariat 7 Accounting
Secretariat 8 Sponsorship and
Exhibition
Secretariat 9 Social Program
/Meals
Secretariat 3 Marketing and
Advertising



∆ milestone
∆ ∆ ∆
Figure 4.10 Example upper level program (simplified); example project involving a
conference. (Work parcel names have been abbreviated because of space
limitations – work items are implied.)
The programs at the higher levels are aggregations or integrations of
programs at lower levels. The lower levels are equivalent to subprojects,
sub-subprojects, . . . Resource, cost and duration information at the lower
levels is folded up to higher levels (Figure 4.11). The better network analysis
computer packages cater for different levels of reporting.
Program time horizons
For communicating the outcomes of planning, for any significant project,
there is commonly a need for a number of programs covering different
time horizons and associated different degrees of detail. There is no con-
sensus as to the number of program types or the time horizon selected for
each. Figure 4.12 shows four program types and a range of common time
horizons, but different representations may be adopted depending on the sit-
uation. Figure 4.12 is based on the assumption that early in the project and
throughout the project there will be a need for an overall project program
(longest time horizon program) showing major work parcels, milestones
and key resources.
Upper level
Property acquisition
Design
Approvals Construction
Commissioning/handover
Intermediate level
Excavation
Footings
Columns, gr–1st fl.
etc. 10–20 broad activities
Slab, beams 1st fl.
etc. 100–200 activities
Lower level
Falsework
Formwork
Reinforcement
Concrete
Curing
etc. 1000 activities
Figure 4.11 Different levels of programs; example, construction of a multistorey
building. (Work parcel/activity names have been abbreviated because of
space limitations – work items are implied.)
Overall project
program
Medium range
program
Short range
program
Daily program
Time
0
0
0
0
Tracks delays, work method changes,
scope changes
Detail for crews, subcontractors, trades
Immediate work effort
Major work parcels, milestones,
key resources
Figure 4.12 Example programs with different time horizons.
The planning process 143
As the project progresses there will be a need for a medium-range program
showing detail over the next, say, 2–3 months and updated, say, monthly.
Updates will take account of, for example, delays, work method changes
and scope changes.
At the workface level for individual crews, subcontractors and trades-
people there will be a need for a short range program showing detail over
the next, say, 2–3 weeks and updated, say, weekly.
There is also a place for a daily program (shortest time horizon program)
showing immediate work effort. This program need be no more than a list
of items.
Where the project is being fast-tracked, and to ensure that the project
proceeds speedily, a starter program may be issued along with the prelimi-
nary project program. The starter program lists the important activities to
be performed during the first, say, 30–90 days of the project while other
programming is being developed.
Detailed execution (fabrication, construction, . . . ) programs are devel-
oped during the design phase prior to the start of any physical work.
The above time horizons are examples only. For any given project, the
time horizons are selected to suit.
Summary
Table 4.2 attempts to summarise the various types of programs used on
projects. The levels correspond roughly with the levels in an organisational
structure. There is no consensus as to the usage of terminology in Table 4.2,
or to the number of rows in Table 4.2.
Some organisations go to the trouble of creating a summary document,
describing all programs, when they are to be produced during project exe-
cution, what they must contain, and to whom they will be directed.
Table 4.2 Example summary program forms
Level Coverage Extent or scale Unit Degree of detail End-user
Upper Outline Entire project Month Low Management
Outline/broad Project design;
Execution/
implementation
Week Low/medium
Middle Execution/
implementation
Execution/
implementation
period
Week Medium Project office
Short term 5–10 weeks Day Medium/high
Lower Weekly 1–2 weeks Half-day High Workface
Daily 1–2 days Hour High
144 Conventional treatment from a systems perspective
Case example – Aerospace industry
This case example discusses the use of schedules and network dia-
grams used within an aerospace company for the management and
reporting of progress on its projects.
The master schedules formed part of a contract with the respective
customer, monitored by both parties at formal review meetings with
actions required if the project falls behind schedule.
Resource planning and cost impact were not performed adequately
until the use of PC-based planning packages. Formerly, schedules
were shown as both broad hand-drawn bar charts and cumulative
production plots (linear schedules). The packages have enabled more
detail to be included in the planning.
The master schedules tended to only show major events, or mile-
stones, and total duration. For example, the manufacture of 200 metal
components for use in final assembly was shown as one bar, not 200
bars, with a total duration. Possible inclusions within the schedule
included:

Despatch date

Paint start and finish

Final assembly

Carbon fibre component start and finish

Metal component manufacture – in batches

Raw material requirements, usually shown as a milestone.
Cumulativeproductionplots werealsousedinbroadform(Figure4.13).
Because each completed component was expensive, the above inclu-
sions were repeated for every product despatched over a two- to
three-year period.
A broad network diagram, hand drawn, was put together before the
first master schedule was completed. The activity durations, and
the logic of the network diagram, were reviewed continually to reflect
the true manufacturing situation.
During the implementation phase, the (broad) network diagram
and (broad) master schedule were changed to reflect the new situation
(ahead/behind schedule) and when the company would be back on
schedule.
The introduction of PC-based scheduling packages bought about
changes (more detail) in the presentation of master schedules.
The company’s planning practices included liaising with the cus-
tomer in establishing milestones, liaising with production as to real-
istic durations, and monitoring actual progress against as-planned.
The planning process 145
Time
Paint and despatch
Batch manufacture
of metal components
and procurement
of raw materials
U
n
i
t
s
Final assembly
Carbon fibre manufacture
Figure 4.13 Example cumulative production plot. Units are the quantity
produced during the contract period or part thereof.
Formerly, without the critical path being known, the people respon-
sible for performing the actual work were not aware of the importance
in adhering to designated schedules. This also created a problem for
the project manager.
Where progress fell behind schedule, recovery schedules were
derived to bring the overall project back into line. Under these circum-
stances, each aspect of the network diagram was reviewed, relevant
people were asked to reassess the activities and when they could be
started/completed. This would lead, amongst other things, to over-
lapping activities.
Case example – Water treatment plant
The company in this case example is a designer, manufacturer, sup-
plier, installer and operator of engineered solutions and equipment for
the water and wastewater industry. The company has projects of var-
ious sizes and complexity and in various stages of completion under-
way at any one time, and spread geographically. This case example
focuses on a project to develop a water treatment plant for a mine.
The company was awarded a contract by the mine for the design,
supply, installation and commissioning of a water treatment plant to
146 Conventional treatment from a systems perspective
supply demineralised quality water. The contract required a comple-
tion date of one year; there were no separable portions.
As part of its tender submission the company provided a program
in simple bar chart format setting out the fundamental elements and
work order of the project. This program contained some thirty ele-
ments without background assumptions.
Immediately upon contract award, the company provided a very
detailed program. A work breakdown structure gave the fundamental
elements of the work. The program consisted of some 400 activities,
and identified the critical path. An S curve (money) was also included.
Two months after starting work, by negotiated variation, the mine
increased the output of the plant. This required a revised process and
alternative equipment to provide for the increased flows. As a result of
the process change, the project was split and two separable portions
were formed with staggered handover dates.
A revised program detailing the new process and methodology was
submitted, this time with some 500 activities. Again the critical path
was identified and an S curve (money) given.
Under the terms of the contract, a detailed monthly progress report
was to be submitted as an ongoing task. Each report contained an
updated program with progress reported against activities and the
S curve, together with derivative programs such as four-week look
ahead, engineering and procurement detail programs.
As the project unfolded, many special in-house subprograms were
produced as working tools. Procurement programs for items such as
plant valves were developed. The procurement of the project valves
as a total group was identified on the subprogram as engineer, size,
enquiry, select, order, manufacture and delivery work without regard
or distinction between the types, makes and manufacturers. The sub-
programcatered for all the variables and delivery of the selected valves
to the site, as and when required.
Exercises
1. Programs are usually prepared with different degrees of detail for dif-
ferent end-users. For example top management requires broad overview
information, while people at the workface need very detailed task-by-task
information.
Consider a project with which you have been involved and consider
the number of levels of people involved. How many programs would be
necessary and what degree of detail would go into each program?
The planning process 147
2. Rank, in your opinion, the following poor practices in programs:

Too much or too little information

Development without foresight as to the program end-users


Lack of continuous updating

The use of unsuitable computer packages

Lack of identifying programs as part of contract administration
procedures

Lack of support of programs with planning information (assumptions
on methods, resources, constraints, . . . )

Completion times given based on inadequate data.
[

Program end-users refer to those who are to use the plan, the so-called
‘doers’.]
Add others to this list, that you may be aware of, and also rank all the
practices that, in your opinion, cause the greatest detriment through to the
least detriment.
What are the causes of these practices? A cause–effect diagram might be
helpful in answering this question.
What suggestions would you make to eliminate the practices?
3. For a project with which you are familiar, what would be the most
important tasks that should be incorporated in a preliminary project pro-
gram/schedule?
4. Consider a project with which you have been involved and consider
the time horizons of the programs involved. How many programs were
necessary and what time horizons were used in each program?
Chapter 5
Planning over levels
Introduction
Simplifications
Planning is simplified in a number of pragmatic ways, one of which is over
levels (Figure 5.1):

The project (work) is broken down (hierarchically) to lower level
subprojects, sub-subprojects and so on to activities, elements and
constituents. Subprojects may be chosen such that the interaction
between subprojects is minimal and is commonly neglected, permitting
subprojects to be planned independently of other subprojects. The same
holds true where multiple projects are combined on an organisation
level. However, interaction at levels lower than subproject is commonly
strong.
Interaction occurs because of the work method (implying an
order/sequencing of work). Iterations (between levels) can largely be
removed by the planner being aware of what is happening on other lev-
els when solving an individual level problem. The levels are uncoupled
as far as possible, for tractability reasons, to make the level problems
less complicated.
Planning goes from the general to the particular, from coarse to fine
(broad to detailed) as the levels descend. Detail emerges from lower levels.
Work breakdown
The work in a project may be systematically broken down to subprojects
to sub-subprojects and so on to component activities. Such a work break-
down can be represented as a tree-type diagram called a work breakdown
structure (WBS). The WBS, however, says nothing of interaction (or inter-
relationship) between subprojects, between sub-subprojects, . . . , between
Planning over levels 149
Planning problem
simplifications
Conversion from
synthesis to
iterative analysis
Hierarchical
breakdown of
a project
Conversion from
probabilistic to
deterministic
Figure 5.1 Simplifications to make the planning problem more tractable.
activities. A work breakdown also says nothing about events, though loose
practice by many writers confuses this.
A systematic breakdown is preferred over an ad hoc approach because
of the increased likelihood of identifying all activities.
Hierarchical modelling
Levels
A model of a project for planning purposes can be built from the bottom-up
using component parts. Figure 5.2 shows two possible representations. Con-
stituents feed up to elements; elements combine to give an activity; activities
combine to give a project (and projects can be combined to give a ‘pro-
gram’). In Figure 5.2, subproject- and sub-subproject-type levels (equiva-
lently work packages) can be inserted, but add nothing new to the following
development; a sub-subproject and a subproject (and work packages) are
no different in character to a project. Similarly, inserting a ‘program’ level
adds nothing new; a program is no different in character to a project.
Formalism
Chapter 9 develops formal project models.
Control
The following examples of controls or actions are not totally exhaustive.
Controls chosen (actions that are taken by the planner) depend on the
planner’s experience and expertise, as well as the situation and a bit of
creativity. Some choices may be better than others, but generally planners
are after a satisfactory solution rather than some theoretically optimum
solution (within an iterative analysis approach to planning).
Controls at any level contain lower level control information.
150 Conventional treatment from a systems perspective
Project
Activity
Element
Constituent
(a)
Activities
Elements
Constituents
Activity
model
Activity
interaction
Element
model
Element
interaction
Project
level
Activity
level
Element
level
Constituent
model
Constituent
interaction
Constituent
level
Constituent
model
(b)
Figure 5.2 Building up a project model.
Project level controls
Controls at the project level relate to the work method perhaps involving
the interaction of activities (and this implies an order/sequence of work).
Examples include:

Examining alternative methods of work. This will lead to alternative
network logic. Examples include outsourcing or contracting out some
work instead of doing the work in-house, different techniques, differ-
ent ways of conducting activities, different work sequences or logical
ordering of activities, different managerial and technical approaches to
the work, using machinery instead of labour, and prefabrication rather
than an in situ work method.

Using overlapping relationships between activities. In place of more
usual finish-to-start (F/S) relationships, start-to-start (S/S), finish-
to-finish (F/F), or start-to-finish (S/F) relationships might be possible,
as well as finish-to-start (F/S) relationships with lead time periods.

Run concurrent/parallel activities. Some activities can be carried out
sequentially, some simultaneously.
Planning over levels 151

Expediting materials; introducing ‘just-in-time’ (JIT) or other inventory
practices.

Extending the project duration. All activities then become non-critical.
The simpler planning problem fixes the work method and only leaves
resource usage and resource production rates as the variables to be deter-
mined. That is, the controls at the project level are fixed, and only lower
level controls remain to be selected.
Activity level controls
At the activity level, control relates to resource usage within an activity:

Total quantity – Use additional/less resource numbers, for each resource
type.

Distribution in time – Compress activities, lengthen activities and split
activities.
In conventional planning parlance, this is achieved through, for example:

Adjustment of activity resource numbers (each resource type) – dis-
tribution over the activity, decreased quantity or increased quantity.
Different ways of using resources (or money) on activities. This may go
under the name of resource redeployment or redistribution. Adjusting
resources may lead to changed activity durations.

Compression of activities. Compressing the durations of activities
involves shortening the activity durations; this usually comes with
an additional and a more concentrated resource requirement (and an
increased cost).

Lengthening activities, though this may come with lower production
and a cost penalty, sometimes called uneconomical drag out.

Splitting activities that can be split. Splitting implies starting an activity,
stopping the activity, starting the activity after a break and so on.

Making use of any activity float. Typically the start dates of non-
critical activities are moved. Note that shifting the start dates of critical
activities will extend the project completion date.

Changing the work calendars.
Element level controls
At the element level, there is a collection of resources (different resources
and possibly different numbers of each resource type). Control relates to the
selection of the number (quantity) and type of each resource. For example,
152 Conventional treatment from a systems perspective
working multiple shifts or overtime fits within this meaning. Different num-
bers and types of resources lead to different production.
Resource information can be collectively measured in a money unit.
Constituent level controls
The constituent level is the basic behaviour level of a resource (person, or
piece of equipment). As such, the controls are the resource production rates
or equivalents of each resource type. Resource production rates, for example
for people, can be altered by the use of a carrot (incentive, reward) or stick,
changing the work conditions, having the correct tools and specialisation.
Resource production rates are different for different pieces of equipment.
Work breakdown
General
Work breakdown is a top-down approach, as opposed to the hierarchical
modelling, given in the previous sections and as discussed later, which is
developed from the bottom-up.
A project (scope) can be decomposed into subprojects and then into
sub-subprojects and so on down to the activity level (but usually not finer).
The lower levels represent finer detail. The process of decomposing a project
might be called scope delineation (or scope definition) by some writers. The
term ‘work breakdown structure’ (WBS) is used to refer to the resulting
systematic decomposition (as opposed to an unstructured activity list where
there is a possibility that some activities have been overlooked).
What is not evident from the work breakdown structure is the degree
of interrelationship between its components, or the reliance of some on
others, and how the work breakdown structure has been used to separate
individual components.
Two wrong and confusing practices that can be seen with some writ-
ers are:

The inclusion of events (and end-products and ‘deliverables’) in a work
breakdown. A work breakdown says nothing about events. An event is
the start or end of a work item; it is not work. The systematic nature
of a work breakdown is destroyed by the ad hoc introduction of events
and deliverables. This is not to say that events and deliverables are
not important for project management purposes, but rather events and
deliverables have no place in a work breakdown; as well arbitrary or
ad hoc practices are to be avoided.

The labelling of entries in the work breakdown as non-work items. For
example, a work breakdown entry may be labelled as ‘book’ when it
should be ‘some work involving the book’.
Planning over levels 153
The work breakdown structure provides a common framework for dealing
with all aspects of planning a project (in an iterative analysis sense) – a
divide-and-conquer approach. It enables a clear understanding of the scope,
and reduces the likelihood that activities will go unthought of (thereby
avoiding disruption to the project rhythm, the incurring of extra dura-
tion and resources and hence extra cost, and the consequent effects on
team morale).
Representation
Work breakdown may be represented as a tree diagram (Figure 5.3),
or appropriately indented dot points, appropriately numbered items
(Figure 5.4) that recognise the hierarchy involved, or as in Figure 5.5.
With respect to Figure 5.4, there is no set standard for numbering
schemes, or numbering for accounting purposes. Rather, the schemes may
reflect different project characteristics and different organisational ways of
dealing, for example, with money.
Example
Consider a project involving the staging of a conference. Major subprojects
would centre on work associated with:

Venue

Printing and publications

Marketing and advertising

Audio-visual equipment
Sub-subproject
Project
Subproject
.
.
.
Activity
Figure 5.3 Work breakdown structure (tree diagram form).
154 Conventional treatment from a systems perspective
Project
1000 Subproject
1100 Sub-subproject
1101 Activity
1102 Activity
1103 etc.
1200 Sub-subproject
1201 Activity
1202 etc.
1300 etc.
2000 Subproject
3000 etc.
Figure 5.4 Example hierarchical breakdown and numbering of activities.
Project
Subproject/
sub-subproject/
work package/
jobs/
subjobs
Activities
Figure 5.5 Hierarchical decomposition of a project.

Registration

Speakers/guests

Accounting

Sponsorship and exhibition

Social program/meals.
These may be further subdivided into their component parts. Figure 5.6
shows how this may be organised hierarchically in a tree diagram. The tree
Planning over levels 155
Conference
project
Printing,
publications
work
Venue
work
Hire
activities
Security
activities
Display,
decoration
activities
Marketing,
advertising
work
Figure 5.6 Tree diagram conference example.
diagram (Figure 5.6) says nothing about the order of and the interaction
among the work.
Subproject choice
Subprojects may be chosen at the convenience of the planner but might
follow a breakdown according to:

Phase (chronological); time period phasing

Work type or nature; resource groupings

Work parcels; the way the work is to be carried out

Contracts

Region, location or geographical area

Organisational breakdown

(Existing) cost codes, cost centres, cost headings; project cost and data
summary reports

The nature of the planning network; activity groupings

Function.
The particular subdivisions chosen in the breakdown will reflect some con-
venient way of parcelling and doing the work. Different organisations and
people, if given the same project, may choose to subdivide it differently.
It can be helpful to list the subprojects in the order in which they might
normally be expected to occur on the project, but this is not necessary.
Irrespective of the breakdown adopted, the collective bottom level of
activities should be the same. All the work that has to be done to complete
a project is contained within the work breakdown structure.
156 Conventional treatment from a systems perspective
Commercially available computer packages used to assist planning com-
monly adopt a work breakdown structure as a basis for data input.
Lower levels can be integrated or aggregated (folded up) to higher levels
for reporting or communication purposes. Any level is the sum of the
immediately lower level work.
With this aggregation, comes an aggregation of duration, resource and
cost information, such that information on durations, resources and cost
can be given at each level. Monitoring (project production/progress and
resource usage), reporting (planned and actual production/progress and
resource usage comparison) and replanning can be carried out at each
level.
Number of levels
The number of levels needed to describe all the project work might range
from about two to about five. The lowest level is chosen to give man-
ageable size entities or identifiable pieces of work; the activities are able
to be characterised adequately in terms of duration, resource and cost
estimates; responsibility and authority can be assigned to each. Until the
breakdown reaches the level of activities, it is difficult to undertake estimat-
ing accurately. Resourcing, costing, scheduling, uncertainties, progressive
measurement and reporting relate to these activities and estimates. What is
‘adequate’ for estimating purposes may change over the course of a project.
The breakdown may evolve as the project progresses, as the project becomes
more definite.
Different stakeholders may choose to decompose to different levels. For
example, a project manager may stop at the level of trade contract packages,
while each trade contractor will subdivide its work package to smaller
items.
Complete decomposition may not occur in some instances. While certain
areas may be well defined and broken down into clear and measurable
components, others may be defined only to a level where a general under-
standing of the activities is evident. This may be due to the complicated
nature of these components or further breakdown may be considered unnec-
essary. This lack of complete breakdown may lead to some problems later
in the project with some planning and design information being found to
be incomplete.
In practice, the development of a work breakdown structure may proceed
simultaneously from the top-down and from the bottom-up.
At the lowest level of this hierarchical breakdown of a project the enti-
ties are manageable components, namely individual activities. These are of
relatively short time span. At intermediate levels the entities are of longer
time span.
Planning over levels 157
Divisional responsibility
Departmental responsibility
Project
Subprojects
Organisation
Work packages
Project
manager
Project team
Individual responsibility Activities
Figure 5.7 Possible relationship between work breakdown and organisational
breakdown.
Einstein is said to have remarked that: Everything should be made as
simple as possible, but not simpler.
WBS relationship to OBS
The project breakdown may bear a close relationship with organisational
breakdown (Figure 5.7) – project team structure when applied to a project,
or organisational breakdown structure (OBS) when applied to an organi-
sation and its functional lines – a representation of the personnel involved
and their interrelationships.
Change
Change may be viewed as:

Changes to the nature of activities within the work breakdown structure
and

Required additional/less activities.
The second category represents scope changes. Changes to the nature
of activities within the work breakdown structure may not normally be
regarded as scope changes, though from a contractual viewpoint might be
regarded as a variation.
Example – Writing a book
Consider the writing of a book involving the researching of numerous
sources, and contributions by different authors. A breakdown of the work
158 Conventional treatment from a systems perspective
etc. writing paragraphs, sentences, words, letters
x.00
x.x0
x.xx
Writing etc.
book
Writing etc.
chapter
Writing etc.
section
Writing etc.
subsection
Figure 5.8 Example WBS for compiling a book.
might be viewed according to Figure 5.8. Lower subdivisions to writing
paragraphs, sentences, words and letters could be considered but are per-
haps too small to be manageable.
The development of the WBS proceeds from the top-down but relies on
the end-users (writers) providing input of what is required at the lower
levels.
Cost account numbers could parallel the chapter/section/subsection
numbering scheme. A responsibility matrix could be set up identi-
fying each chapter/section/subsection with a writer, checker and edi-
tor. Duration, resource and cost estimates for the writing of each
chapter/section/subsection can be made, aggregated to give an overall dura-
tion, resource and cost estimate for the book, and a schedule and budget
for its writing. Monitoring and replanning the writing of the book can be
directly related to chapters/sections/subsections.
Example – Building a house
Consider the building of a low-cost house. The component activities might
be grouped as in Figure 5.9.
Figure 5.10 shows schematically this breakdown.
Example – Road project
A breakdown for a road project might look something like Figure 5.11.
1 Preparatory/preliminary work
1.1 Erection of site facilities
1.2 Survey and layout site
2 Foundation work
2.1 Excavation
2.2 Setting formwork and reinforcing steel
2.3 Concreting
3 Floor construction
3.1 Setting formwork and reinforcing steel
3.2 Concreting
4 Wall construction
4.1 Bricklaying
4.2 Installation of door and window frames
4.3 Plastering
4.4 Painting
5 Roof/ceiling construction
5.1 Roof framing
5.2 Placing roof tiles
5.3 Installing ceiling
6 Plumbing/electrical work
6.1 Excavation, install septic tank
6.2 Laying drainage, sewerage pipes
6.3 Laying water pipes
6.4 Running electric wiring
7 Other work
7.1 Laying driveway, footpath
7.2 Gardening
7.3 Cleaning up
Figure 5.9 Example work breakdown for the building of a low-cost house.
1 2 4 6
1.1
1.2
2.1
2.2
2.3
3.1
3.2
4.1
4.2
4.3
4.4
5.1
5.2
5.3
6.1
6.2
6.3
6.4
7.1
7.2
7.3
House
building
3 5 7
Figure 5.10 Example work breakdown for the building of a low-cost house.
160 Conventional treatment from a systems perspective
Road
project
Earthworks
Excavation
Drill, blast
Transport
Fill
Survey
work
Base, seal
work
Culverts,
bridges
work
Figure 5.11 WBS for an example road project.
Case example – Truck research
This case example describes the work breakdown structure for a
project involving the examination of the operation of mining haul
trucks, from the viewpoint of the owner of the mine and the trucks.
Research work is implied, but not stated, in the breakdown that
follows.
The case example relates to a research and development project
established due to inadequate brake, transmission and differential life
of mining haul trucks. The overall aim from the owner’s view was
to reduce costs associated with the operation of the haul trucks. The
mine looking after these trucks had to take the research results and
address them wherever practical. After preliminary research into the
trucks, it was noticed that the majority of costs incurred were in
maintenance, and in particular relating to brakes, transmission and
differential life. The mine was uncertain as to why these trends were
occurring and hence was required to do further research. A research
and development project was established.
Due to the expansive nature of the project, it was first divided into
research work associated with three subprojects:
1. Issues relating to brake life
2. Issues relating to transmission life
3. Issues relating to differential life.
Procedures used and decisions made, up to and including this (first)
division, all constituted initial stages of planning. A problem was
identified and a project developed to investigate that problem. There
were many issues which affected the costs of haul trucks; honing in on
Planning over levels 161
maintenance issues already created a subproject, because there were
other ways that costs on the haul trucks could be addressed. The
mine had effectively narrowed down an area based on collected data,
where they thought that significant cost savings could be produced.
However, if this path was proved to be wrong, then they needed to
track back to this first division and follow a different path.
Once the overall project had been given a definite subproject to
follow, this was further broken down into three distinct areas based
on truck components. At this point, creating subprojects helped to
keep the overall project’s aims in sight whilst breaking the project
into subprojects and sub-subprojects enabled easier management and
planning. If work breakdowns had not occurred, through a systematic
approach backed up by previously collated data, then identifying an
area where most expense on the trucks was occurring would have
been a burdensome task. It also followed, that had the first division
not been backed up by history files, experience etc., honing in on the
correct components that were of the greatest cost would not have been
possible. Not much effort was spent looking into the detail of why or
how the problems were occurring, but rather what was costing. The
‘why’ details came forth later in the project breakdown.
Narrowing in on the issues relating to brake life, a new set of aims
(sub-aims) was defined, to assist in keeping the overall project’s aims
on track. A sub-aim of this project was to identify the limiting factors
of brake life for the haul trucks. Once this sub-aim was satisfied,
limiting factors and therefore areas with associated cost were to be
identified and were to help contribute to answering the overall aim.
Even though the subproject had been divided into sub-subprojects
to have reached the above stage, it was still too cumbersome to allow
for factual and clear communication of identified problem areas; so it
was once again broken down smaller, that is to sub-sub-subprojects.
There were many factors which may lead to limiting the brake life
of these trucks at this level. Two clearly separate issues relating to
research on brake life were identified (second division):
1.1 Factors relating to the application of the trucks
1.2 Factors relating to the design of the trucks.
Within each of these, a further breakdown was evident; this next
breakdown addressed research work areas that were categorised
according to the above factors (third division):
1.1.1 Speed in relation to brake effectiveness
1.1.2 Haul road profiles and effect on braking system
1.1.3 Loading of vehicles
162 Conventional treatment from a systems perspective
1.1.4 Operator considerations/human issues
1.1.5 Maintenance considerations.
and
1.2.1 Characteristics of fluid in hydraulic system
1.2.2 Material of brake pads
1.2.3 Design characteristic monitoring
1.2.4 Different types of braking systems
1.2.5 Environmental effects
1.2.6 Pressure system and fluid flow for hydraulic brakes.
The second division was made by the subproject manager so that
the sub-aims could be clearly met. This division was based on
brainstorming all possible contributing factors to brake life. After
brainstorming it became evident that two main factors could be con-
tributing to the brake costs. The items which fell into these two main
categories went together to make the third division. The procedure
followed to come to the second and third divisions differed from that
followed to come to the first. This was because these later two divi-
sions were the initial focus on not only what could be contributing
to the brake life but also why these categories could be contributing
to the brake life problems. The second division and the ones that
followed were based on hypothetical possibilities; there was no firm
previous data that could have been collected to give the best direction.
Activities and work procedures were then identified and were com-
bined in addressing the set of aims associated with the third divi-
sion. These were seen as one of the last steps in the breakdown tree.
The work procedures highlighted how the third division aims con-
tributed to the overall project’s aims. Future steps then became specific
activities rather than areas and so on.
Case example – Bridge construction
This case example describes the work breakdown structure (WBS) for
the construction of a bridge from the viewpoint of the contractor.
For the incrementally launched box-girder bridge, the breakdown
chosen stemmed from a combination of planning requirements and
the ease by which the project manager could monitor, manage and
record construction activities and costs associated with each subpro-
ject. General agreement and past experience between the planning
Planning over levels 163
and construction teams limited the number of work breakdown levels.
The work breakdown structure for the project consisted of four main
subprojects, namely work associated with:
1 Substructure
2 Piers and abutments
3 Box girder
4 Completion of superstructure.
Each of these subprojects was further divided into work associated
with:

Plant and equipment
– Hire
– Maintenance

Concrete

Reinforcement
– Subcontract
– Materials

Formwork
– Subcontract
– Materials.
Separate work breakdown structures existed for activities associated
with site mobilisation and demobilisation, electricity, tools, safety,
consumables, site overheads and so on.
The breakdown of the project into these four major subprojects
was intended both for ease of record keeping and for usefulness of
the records kept. Whilst box-girder bridges may have different pil-
ing (substructure) and pier column arrangements, information on the
costs of building that particular type of box girder, regardless of the
foundations, is useful and hence was recorded. Similarly, by separat-
ing the substructure work, records were now available on the costs
and production of the barge, the piling hammer and the manufacture
of pre-stressed concrete piles. This enabled bridge records to be used
for a variety of other projects, for example wharves.
The work breakdown structure also followed the course of the
job. By relating the work breakdown structure to the master plan,
the project manager was able to monitor the project, in both schedule
and cost senses. Schedule management was achieved by being able
to systematically close-out subprojects. Cost savings in a completed
164 Conventional treatment from a systems perspective
subproject could be used to offset the increased costs in another sub-
project, enabling the project to remain within overall budget.
The subsequent breakdown into work associated with plant and
equipment, materials and labour was largely a breakdown based
on costs. The majority of plant and equipment for the project was
obtained on internal hire; thus the breakdown assisted in the prepa-
ration of monthly accounts. Materials were assumed to be estimated
accurately at tender stage, and subsequent blow-outs in this area
pointed to estimating problems. Labour costs, the greatest variable,
gave an indication of the management of the project.
Case example – Computer system upgrade
In this case example project, the work breakdown structure (WBS)
was not clearly understood by the stakeholders, resulting in confusion,
frustration and near disaster.
The project, with its end result of upgraded technology throughout
a firm, involved:

Upgrading all hardware

Upgrading all hubs and associated cabling

Upgrading all operating systems

Updating application software

Training of all staff in the new software applications.
The first three of these were achieved with only relatively minor
troubles and concerns, while the fourth proved to be a major headache
due to the number of subprojects it contained and the necessity to
outsource five of these subprojects to five separate organisations – a
decision based on a lack of required expertise and experience within
the firm.
At the completion of planning, and by the time it came to update
the application software, it was assumed by all stakeholders that
‘all bases had been covered’, the stakeholders understood their roles
and those of others within the project, and nothing had been over-
looked. Various companies were working together on the document
macro and template developments and design of a conversion fil-
ter for existing documents. Other companies were working on the
design of the document management system profiles for the new doc-
uments to be created and their conversion. Everyone, it seemed, was
actively engaged in working towards converting existing documents
Planning over levels 165
and document management system profiles over to their new format,
and preparing for the creation of new documents and profiles.
However, no one had been assigned responsibility for the actual
conversion of the existing word-processing documents. Each stake-
holder assumed that another was responsible for this activity because
everyone had seen and worked with the conversion filter at some stage
and had discussed it at some point during the project. It was assumed
by all, therefore, that preparations for the commencement of training
were in hand.
Fortunately, this error was discovered the day before training was
to commence, soon enough to avert a near disaster, should it have
remained undetected. Training was to be in groups of six staff mem-
bers at a time, offsite, in all aspects of the newsystem. While each group
was in training, several other activities were to occur concurrently: the
rolling out of the new desktops onto their PCs, the conversion of their
word-processing documents and their document management profiles.
Hadtheir documents not beenconvertedupontheir returnfromtrain-
ing, they would not have been able to access critical work. Their doc-
ument profiles would have been converted to the new document man-
agement system and their PCs would have been ready with the new
system but they would not have been able to find any existing work.
It would have been impossible to convert their documents across in
time for them to work for the first few days after training, and because
thousands of these documents were vital legal documents, this was a
situation which would have had serious repercussions. It could have
been easily avoided by better planning and closer attention to detail.
The project had continual changes within each subproject, as the
software was developed.
Subproject work that was done by outside contract lacked a scope
statement defining the extent of the work. A few faxes confirming
verbal quotations were all that were present, but even these did noth-
ing to state the overall area of responsibility or scope (of work) to be
undertaken by each contractor or the firm.
The project manager was appointed from within the firm. Despite
having no prior project management experience, he had the neces-
sary skills to understand the scope (of the project), and was a good
manager, with good people and communication skills. He also had
the support of senior management, who nevertheless expected him to
add it to his normal duties. As a result of this, the person who was
expected to be the main impetus behind the success of the project
was so busy and so much in demand that, instead of being able to
dedicate a worthwhile period of time and attention to ensuring that
each activity was performed, was constantly spreading himself around
in smaller, superficial time lots, without being able to really manage
166 Conventional treatment from a systems perspective
the project. When he requested assistance from senior management it
was refused, with the reason that ‘no more resources (people) could
be spared’. As a result of this limit on human resource numbers, other
problems arose. One of these was that documentation of the progress
of the project was sparse. Perhaps if such progress had been docu-
mented, it would then have drawn attention to the oversight on the
activity of converting the documents.
Documentation could, at a minimum, have included a detailed
WBS, with activities delegated to appropriate members of the team.
It may have drawn attention to the impending problem.
Had the documents not been converted in time, the inevitable inabil-
ity to access work would have occurred for the trainees. This may
have had the effect of causing frustration, accusations and, possibly,
arguments. It also would have caused a blow-out in the schedule and
budget because the firm possibly would have had to then insist that
some of the trainees get involved in the conversion process, thereby
taking them away from other work.
Iterative analysis approach
Generally, planning over levels is practised in conjunction with an iterative
analysis mode of attack, primarily because the synthesis version is too
difficult (Figure 5.12).
The iterative analysis approach follows Figure 5.13, showing the feedback
loops, and in particular the feedback loops between levels. Figure 5.13 is
saying the same thing as Figure 2.9.
The initial solution guess is commonly based on past projects, experience
and knowledge of the industry.
The analysis and evaluation will generally involve network analysis and
the use of time-scaled networks, bar charts, cumulative production plots and
so on. The analysis may be assisted by 2D and 3D computer visualisation,
scaled models and simulation.
Conversion from
probabilistic to
deterministic
Hierarchical
breakdown of
a project
Conversion from
synthesis to
iterative analysis
Planning problem
simplifications
Figure 5.12 Simplifications to make the planning problem more tractable.
Planning over levels 167
Activity resource
selection
Analysis, evaluation
Project
level
Activity
level
Adjustment of
project level
controls
Adjustment of
activity level
controls
Adjustment of
value system
Planning complete
Owner's values
objectives, constraints
Objectives, constraints,
performance –
not satisfactory
Objectives,
constraints,
performance –
satisfactory
Element resource
number, type
selection
Constituent production
rate selection
Work method selection
(inc. scope, work
breakdown, sequencing)
Element
level
Constituent
level
Adjustment of
element level
controls
Adjustment of
constituent level
controls
Figure 5.13 Iterative analysis approach to planning over levels.
Controls selected depend on the planner’s experience and expertise, as
well as the situation and some creativity. Some solutions may be better than
others. Generally planners are after a satisfactory solution rather than some
theoretically optimum solution.
If one approach doesn’t work or doesn’t work completely, then other
approaches may be tried. Approaches are only limited by the planner’s
imagination. Knowledge of the project and the industry will help develop
168 Conventional treatment from a systems perspective
alternative approaches. Whether any of these approaches are possible will
be determined by the project and the nature of the project work.
Any control adjustment, amongst other things, may:

Lead to new cumulative resource (money) plots

Alter the critical paths

Lead to rescheduling of activities.
Value system adjustment may redefine the planning problem or, for exam-
ple, lead to the removal of physical, technical or logic constraints affecting
work.
Example
To assist an understanding of planning over levels, an example is presented.
The example project chosen is one which most people should be able to
visualise. The procedure in planning the project over levels is developed.
The simpler iterative analysis approach (compared to a synthesis
approach) is presented.
End-product and project
An investigation report is required to be produced by a consultant. The
project is the means by which this report comes about.
Objectives and constraints
The project objectives and constraints are from the viewpoint of the con-
sultant undertaking the project (undertaking the work for an owner), but
these could be expected to reflect the owner’s values as well.
Suggested objectives might be:

Minimum cost

Maximum efficiency in work practices, including the usage of
personnel.
Suggested constraints might be:

Availability of expertise and people

Duration upper limit

Cash flow, perhaps requiring payment from the owner as soon as
possible

Fitting in with other projects.
Planning over levels 169
Report
Proposal Literature Field data Hard copy
Initiate
Acceptance
Preliminary
search
Supplementary
search
Analysis
Preliminary
collection
Reduction
Supplementary
collection
Reduction
Analysis
Progress
report
Draft
Final copy
Figure 5.14 Possible work breakdown for the investigation report example. (Activity
names have been abbreviated because of space limitations – work items
are implied.)
Scope
Broad scope
The broad scope (of the project) is one of producing an investigation report.
It involves developing a proposal, literature searches, the collection of field
data and the production of hard copy.
The project proposes to use a number of people and their associated
skills.
Scope delineation – Work breakdown
A possible breakdown to activities, not in any particular order, is given in
Figure 5.14.
Project level
Activity interaction/sequencing
Precedence relationships establish the logic of how the project is to be put
together. The following is one suggested way in which the work might be
carried out.
170 Conventional treatment from a systems perspective
Code Activity Precedence
Prop Initiation of proposal –
Accpt Acceptance of proposal Prop
Prelit Preliminary literature search Accpt
Predat Preliminary data collection Accpt
Prorep Progress report Prelit; Predat
Prered Preliminary data reduction Predat
Suplit Supplementary literature search Prorep
Supdat Supplementary data collection Prorep
Supred Supplementary data reduction Supdat; Prered
Anal Analysis of literature and data Suplit; Supred
Drrep Prepare draft report Anal
Firep Prepare final copy of report Drrep
Network formation
The network for this project can now be drawn. The network may be either
an activity-on-link diagram or an activity-on-node diagram. The network
looks like Figure 5.15(a or b). Figure 5.15a is an activity-on-node diagram.
Figure 5.15b is an equivalent activity-on-link diagram.
Note that the network diagram is no more than the logic of how the
work is to be carried out.
Prop Accpt
Prelit Suplit
Anal Predat
Prered
Supdat
Supred
Prorep Drrep Firep
(a)
Prop Accpt
Prelit
Suplit
Anal
Predat
Prered
Supdat
Supred
Prorep
Drrep Firep
d
Dummy activity
d
(b)
Figure 5.15 Example project networks. (a) Activity-on-node diagram; (b) Activity-on-link
diagram.
Planning over levels 171
Activity level
Activity duration estimates
Activity duration estimates imply that some assumptions have been made
about resourcing and resource production rates. Duration, resourcing and
resource production rates are related.
A possible set of duration estimates for the activities is as follows:
Code Activity Duration (days)
Prop Initiation of proposal 10
Accpt Acceptance of proposal 2
Prelit Preliminary literature search 21
Suplit Supplementary literature search 7
Anal Analysis of literature and data 2+4 = 6
Predat Preliminary data collection 25
Prered Preliminary data reduction 6
Supdat Supplementary data collection 10
Supred Supplementary data reduction 3
Prorep Progress report 5
Drrep Prepare draft report 12
Firep Prepare final copy of report 5
These values obviously depend on the skills of the estimator and the famil-
iarity this person has with the work involved in the project.
Resource estimates
The resource requirements of ‘professionals’ is estimated to be as follows:
Code Activity Professionals/day
Prop Initiation of proposal 2
Accpt Acceptance of proposal 4
Prelit Preliminary literature search 2
Suplit Supplementary literature search 2
Anal Analysis of literature and data 1+1 = 2
Predat Preliminary data collection 3
Prered Preliminary data reduction 2
Supdat Supplementary data collection 3
Supred Supplementary data reduction 2
Prorep Progress report 1
Drrep Prepare draft report 3
Firep Prepare final copy of report 1
172 Conventional treatment from a systems perspective
Money estimates
The following are estimates for costs to undertake the activities:
Code Activity Cost/day ($/day)
Prop Initiation of proposal 1500
Accpt Acceptance of proposal 3000
Prelit Preliminary literature search 1200
Suplit Supplementary literature search 1200
Anal Analysis of literature and data 1200
Predat Preliminary data collection 3000
Prered Preliminary data reduction 1200
Supdat Supplementary data collection 3000
Supred Supplementary data reduction 1200
Prorep Progress report 600
Drrep Prepare draft report 1800
Firep Prepare final copy of report 750
The agreement for payment (income) for the work is that $120 000 be paid
after the completion of the progress report (Prorep), and a further $120 000
be paid on completion of the final report (Firep).
Time-dependent costs have not been included.
Element and constituent levels
The controls at the element and constituent levels are assumed fixed for
this example.
Analysis
Network analysis
See Figure 5.16.
Bar chart
The information from a network analysis is readily transferable into bar
chart and time-scaled network form (Figures 5.17 and 5.18).
Figure 5.17 shows the bar chart plotted for all activities starting at their
earliest start times.
Time-scaled network
Figure 5.18 gives a time-scaled network.
Planning over levels 173
Prop
10
Suplit
7
Anal
6
Prered
6
Drrep
12
Firep
5
42 49
48 55
6
EST EFT
LST LFT
TF
Key :
Duration: days
0 10
0 10
0
10 12
10 12
0
12 33
16 37
4
37 42
37 42
0
12 37
12 37
0
37 43
46 52
9
42 52
42 52
0
52 55
52 55
0
55 61
55 61
0
61 73
61 73
0
73 78
73 78
0
Accpt
2
Predat
25
Prelit
21
Prorep
5
Supdat
10
Supred
3
Figure 5.16 Example network calculations.
10 20 30 40 50 60 80
Prop
Accpt
Prelit
Suplit
Anal
Predat
Supdat
Supred
Prorep
Drrep
Firep
Days 0
Prered
Scheduled duration
Float
Connection
Figure 5.17 Connected bar chart for example project.
Calendar
For use by project personnel, the information is more desirably presented
on a calendar allowing for stipulated work hours per day, weekends, public
holidays, rostered days off (RDO), number of shifts per day and so on.
174 Conventional treatment from a systems perspective
10 20 30 40 50 60 80
Prop
Suplit Anal
Predat
Supdat
Prorep
Drrep Firep
Days
O
O O O
O
O O
O
OO
0
O
Prelit
Scheduled duration
Float or dummy activity
Accpt Prered
Supred
Figure 5.18 Example project time-scaled network.
Resources
Figure 5.19 shows the resource plot for the resource of ‘professionals’.
Money
Plots of money requirements (costs) – expenditure plot, and a cumulative
expenditure plot or S curve – over the duration of the project are given in
Figure 5.20.
With income superimposed, Figure 5.21 results.
10 20 30 40 50 60 80 Days 0
2
4
6
Professionals
Figure 5.19 Example resource plot for project ‘professionals’.
0%
10 20 30 40 50 60 80 Days 0
2
4
6
0
100%
Cumulative
expenditure
Expenditure
$000
Figure 5.20 Example expenditure (upper diagram) and cumulative expenditure
(lower diagram) plots.
10 20
30
40 50 60 80
Days
0
0%
100%
Cumulative
expenditure
Cumulative
income
Figure 5.21 Example cumulative expenditure and cumulative income plots.
176 Conventional treatment from a systems perspective
Evaluation against objectives, constraints and performance
The results of the analysis are now evaluated against the project objec-
tives and constraints. How is the project in terms of schedule/production,
resources and money?
Feedback – Adjustment of controls
The controls are adjusted at the relevant levels based on this evaluation.
For example, should the availability of professionals be fixed at five peo-
ple, then something has to be done to reduce the peak resource requirement.
Conventionally this is called resource constrained scheduling, but is no more
than an adjustment of the controls.
As an another example, the resource usage as depicted in the resource
plot may not be considered efficient. Conventionally resource smoothing
addresses this problem, but again it is no more than an adjustment of the
controls.
There may also be feedback to adjust the project owner’s value system
if the current objectives and constraints are unsuitable in some way. The
objectives and constraints may also be modified if the analysis indicates
something unusual.
Repeat
The whole process is repeated until the objectives and constraint satisfaction
are to the liking of the planner. Experienced planners may only need one,
or even zero, iterations. Inexperienced planners may need many iterations.
Exercises
1. Draw a tree diagram for the assembly of a machine or toy with which
you are familiar. At the top of this diagram is the toy or machine. The next
level represents subsystems; the next again level sub-subsystems and so on.
The bottom level represents the individual nuts and bolts and discrete parts.
No two tree diagrams will be the same. It depends on the machine or toy
and the way it is put together.
2. Assume that you are undertaking a project of holding a technical or
public seminar. Some of the more important activities would involve those
related to:

Attendees (informing, handling queries, enrolling, payment, . . . )

Venue (selection – size, etc., booking, refreshments and meals, payment,
room preparation, audio-visual facilities, whiteboard, . . . )
Planning over levels 177

Speakers (selection, transport, accommodation, fees, . . . )

Proceedings/seminar notes (writing, design, printing, binding, distribu-
tion, certificates, . . . )

Promotion (market establishment, advertising, distribution, . . . ).
All of this has to be done under various schedule/production, resource and
cost constraints.
Develop a work breakdown structure for this project.
Consider now, the way a budget might be developed for this project.
How could the budget format relate to the work breakdown structure?
How do you make duration estimates to carry out each of the earlier
mentioned activities?
Consider now the personnel involved. These might include:

Speakers

Company director

Project manager

Assistants to the project manager

Attendees.
Relate the organisational structure that you might use to the work break-
down structure that you have just developed.
3. How do you know what is the lowest level to decompose a project?
4. A work breakdown structure (WBS) is chosen to suit the needs of each
project. It is an accepted way of representing how a project is broken into
finer components. However, it is possible to not be systematic and describe
the work in an unstructured list.
What is the difference between an unstructured activity list and a WBS?
Why is the latter preferred?
5. Consider a project associated with a social or sporting group to which
you belong. Develop a work breakdown structure for the project.
Draw a responsibility chart/matrix (Chapters 3 and 7) concerning all the
project participants.
Draw an organisational chart identifying all the people involved.
Relate the work breakdown structure to the responsibility chart/matrix
and organisational chart.
Relate duration, resource and cost estimates of the project to the work
breakdown structure.
6. A work breakdown structure subdivides a project into hierarchical
units of subprojects, . . . , work packages, . . . , activities. The end result of
the work breakdown process is a collection of work units or activities of
relatively short duration and of manageable size.
Each activity is of a size to facilitate estimating the number and type of
resources, money and duration that are necessary to accomplish the activity.
178 Conventional treatment from a systems perspective
Each activity is a meaningful job for which individual responsibility can be
assigned.
Consider a project associated with a social or sporting group to which
you belong. What is the lowest level at which estimates of resources, money
and durations can be made?
Chapter 6
Replanning
Introduction
Replanning is carried out according to project stages or at discrete points in
time. Replanning is no different in form to planning. Replanning updates
the controls based on actual performance, the current project state and
some forecast of future influences, as the project progresses. Information
assisting the replanning problem at any stage is better less far into the
future. The (re)planning problem (for the remainder of the project) becomes
better defined as the project progresses. As well, the planner’s experience
and understanding of the project improves. The level problems repeat for
each stage, but with updated information.
During project execution, attention centres on keeping up to date with
the effects of changes, delays and unforeseen events which on many projects
appear inevitable. This implies continual replanning. A good plan allows
for some flexibility and can be adapted to changing circumstances that may
occur during a project. However, attempts are made to foresee potential
influences in upcoming stages, and controls (immediate or contingency)
developed for them much like in risk management (Carmichael, 2004).
Planning starts with broad assumptions which are refined as the
(re)planning progresses. Planning always proceeds from the general to the
detailed. To attempt to plan in great detail right from the beginning ignores
the reality that future events are difficult to predict with certainty and out-
comes never eventuate as intended. As the project progresses, more informa-
tion becomes available, ‘unknowns’ become ‘knowns’, and the (re)planning
can become more detailed and more accurate.
The ability to influence the outcome of a project could be expected to
diminish as the project progresses (something like Figure 6.1). Planning
carried out early in a project accordingly could be expected to provide the
best opportunity to determine the direction, cost and so on for the project.
As the project takes shape, the opportunities disappear. Early, decisions
could be expected to have the greatest effect. Conversely, changes made
later in a project are viewed as more costly; it is considered important to get
it right at the start. In Figure 6.1, trends only are shown and the location of
180 Conventional treatment from a systems perspective
Ability to
influence
cost
Project
start
Project
end
Time
Ability to
influence
cost
High
Low
Cost of
change
Cost of
change
High
Low
Concept Implementation Termination Development
Savings
potential
or
Figure 6.1 Anticipated cost influence and changes over a project’s duration
(Carmichael, 2004).
the intersection point of the trend lines is not important. How data could
be obtained to substantiate the widely held beliefs of Figure 6.1 is unclear
(Carmichael, 2004).
Planning, monitoring, reporting and replanning
Planning establishes a project ‘baseline’ or ‘target’ performance. Disturbance
takes the project behaviour away from the current baseline.
As one precursor to replanning, the current output (performance, state)
of a project is monitored and the information obtained converted into
reports on which management decisions (selection of controls) can be made
(Figure 6.2).
Monitoring involves measuring or observing the output or performance
of a project.
Reporting is in absolute terms and terms relative to planned baseline or
target performance (variances – the differences between planned and actual
performance). Like programs, reporting is made appropriate to the level of
management concerned. Information desirably is presented in a simplified
and easily understood form. Reporting specifics can take various forms.
Replanning incorporates updated estimates of future influences, and uses
the current project state as initial conditions. Variances are used to guide
the choice of which controls need to be adjusted to correct the path of
the project. Replanning gives an updated project baseline or target per-
formance. Disturbance takes the project behaviour away from the current
baseline. Monitoring and reporting follow, and the cycle continues through-
out the project.
Replanning 181
Implement
Monitor Report
Plan/
Replan
Absolute and comparative
performance
Controls
Future
influences
estimate Disturbance
Figure 6.2 Plan–implement–monitor–report cycle.
Generally the greater the frequency of monitoring, it could be expected
that the project will more likely perform close to that desired. Continual
monitoring and feedback of project progress is necessary in a changing
project and environment (reflected by disturbance), else replanning may
reduce to something akin to crisis management. (Project) scope also rarely
stays constant.
Planned-for project states may not be reached because of delays, weather
disruptions, industrial disruptions, variations (design changes, owner pri-
ority changes, . . . ), resource variability (within and between activities) and
so on, even though allowances for these matters may have been included
in the original planning. All these things are collectively referred to as
disturbance.
In principle, replanning is no different to planning. It is planning carried
out subsequent to any original planning. As such, all the discussion on
planning applies to replanning.
Systematic, disciplined replanning is recommended practice.
Performance measures
Performance measures are gauges by which the success or otherwise of a
project is measured. A performance measure is in the lay usage sense of the
word ‘objective’ (mentioned in Carmichael, 2004).
Generally a project has a planned-for performance (resource usage and
production over time; this may be translated to cost), and actual perfor-
mance is compared with planned performance. One measure of a project’s
success may be in terms of the difference between this planned and actual
182 Conventional treatment from a systems perspective
performance. (But comparisons between actual performance and other
performances, for example industry standards/benchmarks, competitors’
performance or previous projects, are possible.)
Performance measures tend to be developed later in projects (after the
initial planning has been done); objectives are developed very early.
A performance measure might be termed a ‘key performance indicator’
(KPI) or similar, but all of these types of terms are used very loosely by
most people. The terms are used frequently by project personnel because
they sound good and impress, but lack precision.
Case example – Powerline construction
This case example looks at compression practices to meet a pow-
erline construction project’s deadline, when factors external to the
project, such as long delays in obtaining government approvals, and
unexpected wet weather arose.
This project started with a route selection study, environmental
impact statement (EIS), design, materials procurement, survey and the
letting of the construction contract. Governmental approval was not
obtained until a year later, which meant that there were two months
less available than originally estimated to construct the powerline,
weather permitting.
Crucial to the success of the project would be close and constant
monitoring of the construction activities to ensure that the project
deadline was met.
The tender documents for the line construction were distributed
just prior to obtaining the government approval to proceed with the
project. So the owner was able to interview the final contenders and
advise them of the project’s requirements, with emphasis being placed
on the project’s deadline and their proposed work schedules to com-
plete the work as required. The actual construction was split into two
parts, a section of the line on a mining lease area, and a section off
the mining lease area.
The contractor was given the responsibility for arranging and coor-
dinating the delivery of the concrete poles with the pole supplier. One
of the major uncertainties of the whole project was the delivery of the
concrete poles to site, in this case to the pre-drilled hole sites, because
of the terrain.
At the same time, other infrastructure work for the mine was being
conducted in the vicinity of the line, namely the laying of a pipeline in
a main mine access road reserve. The road was also the main access
route for the powerline construction, and this made delivery of the
Replanning 183
poles very awkward in the majority of instances. This was an item
which was not known about or considered in the original project plan.
This route was also the main transportation route for the delivery of
equipment to and from the mine site, and this ensured this particular
road was always busy.
The successful contractor established itself on site very promptly.
The first month or so went very smoothly, except for obtaining some
environmental approvals for benching work on the mine site, but this
was quickly cleared up at one of the regular weekly meetings with the
owner.
Then Murphy’s Law came into play; it started to rain. As a
result of this, the poles could not be delivered to site. This meant
that poles were being dumped at specified sites, and additional
cranage and transportation was required. The project started to
fall behind schedule, and all stakeholders were starting to become
concerned.
In conjunction with the contractor, the owner started to look at
what could be done to obtain some slack in the program to ensure
the project would meet the required deadline.
The following items were identified as areas where the project could
be progressed or compressed:

The contractor’s original proposed work schedule was looked at,
and it was discovered that it was one work team short. This was
due to the fact that it had not started pipe stringing as yet. This
was rectified and the additional team was put in place.

Having looked at the proposed schedule, it was mutually decided
to vary the schedule and start work in those areas that were
accessible even though some other planned areas were inaccessible
due to the wet weather. This meant some staffing re-arrangements,
but it seemed to work effectively.

Lengthened work hours were suggested. However, this was
thought to be a major problem, because part of the government
approval conditions stipulated the working hours and days – Sun-
days were not included and the work hours were to be between
7.00 am and 5.00 pm.
Approval for extended working hours was obtained. This
meant the contractor could utilise the additional hours, and the
contractor was also allowed to operate on Sundays provided this
was not done in the vicinity of the local town.
The concern was for the impact of all these items on the estimated
project cost. The other major concern was that the contractor had
184 Conventional treatment from a systems perspective
about two weeks of delays due to wet weather, which could have
meant that the deadline was not going to be met anyway.
The canvassing of options was handled openly among all parties
concerned. This enabled timely and efficient solutions to be iden-
tified and allowed processes to be put in place to resolve these
issues.
The additional cost to the owner was negligible compared to the
cost of not meeting the deadline, and so the contractor was given the
go ahead to work the extended hours.
The contractor still struggled with the resources that it had and
inevitably towards the end, additional resources were required to
ensure that the project deadline was met. In the end, the project
just met the deadline, due in part to good luck rather than good
management. The owner was satisfied and the additional costs were
negligible compared to the overall cost of the project.
Case example – Documentation for lng plant expansion
This case example involves the selection of controls in a project involv-
ing the development of documentation related to the expansion of
a liquefied natural gas (lng) plant. The same contractor who had
successfully designed and built the original plant undertook this doc-
umentation project. The owner had a prominent involvement in the
project, electing to have an owner’s team reside in the same building
as the contractor’s team. This was aimed at improving communica-
tion between the contractor and the owner, as well as to increase the
efficiency of the approval process. The project was initially on a tight
schedule and ran in parallel with finding purchasers of the product
from the expanded plant.
There were many sub-teams involved in the project, each special-
ising in an area of expertise such as electrical and instrumentation,
fabricated equipment, piping, special equipment, rotating equipment
and process engineering. A number of technical groups were com-
bined to form a larger group under the direction of one project engi-
neer. The fabricated equipment group, special equipment group and
rotating equipment group were all under the direction of the same
project engineer. It was the project engineer’s responsibility to record
the progress made by his particular technical groups each week and
report this to the planners.
At the beginning of the project, one of the first tasks that the lead
engineers of each technical group was required to perform was to
Replanning 185
give estimates of hours to finish each document. (The [project] scope
was to produce a set of documents that would detail the require-
ments of the expanded plant.) This information was charted as an
S curve, which formed the basis of all monitoring, reporting and
replanning activities. The hours were approved by the owner and
allocated to the technical groups. Each week the project engineers
would report the progress of their groups to the planners and the
hours expended would be plotted on an S curve so that the dif-
ference between planned and actual hours could be seen. At one
stage, the rotating equipment group had fallen behind schedule (but
had not expended the budget hours to get to the stage they had
reached), while the special equipment group was ahead of schedule
and hours expended. In order for the rotating equipment group to
get back on track, they needed more resources. The owner, how-
ever, had a limit on the resources to be used on the project, and
so to eliminate the need for additional resources, the project engi-
neer rearranged resources for a short period. One engineer from the
special equipment group moved to the rotating equipment group for
one week and performed tasks that relieved that group’s workload.
The rotating equipment group was able to move back onto schedule
without employing additional personnel. It was important, however,
to ensure that all hours worked by the special equipment engineer
on the rotating equipment tasks were allocated to the correct cost
code. For costs to have gone to the special equipment budget would
have been a misrepresentation of actual costs spent in each area,
and hence further planning would have been incorrect due to false
information.
The planners would track progress each week through planning
meetings with the project engineers, where verbal reports on progress
were given. A group being behind schedule, however, was not
the only problem to them. A group being ahead of schedule was
just as troublesome to the planners. The reasons for this were
twofold:

The schedule was based on a flow of information. The special
equipment group relied on information from process engineers.
The process group had a schedule to deliver technical information
and if the special equipment group was ahead of schedule, there
would be a wait involved for technical information from the
process group.

The owner would review the schedule each week. If many groups
were seen to be ahead of schedule and others behind schedule,
it would show poor planning and a poor allocation of resources.
186 Conventional treatment from a systems perspective
This would reflect adversely on the contractor as a professional
engineering consulting firm.
The planners would give the project engineer instructions on where
each group needed to progress in the coming weeks, and then the
project engineer was responsible for reallocating personnel to ensure
that all groups stayed on schedule.
A standard template, for every document, determined the quality of
the documents. This, however, did not exist at the start of the project
but was established with input from the groups that were first to reach
the stage of requiring a template. This showed signs of poor planning
because many groups were held up at the beginning by first having
to create the standard template for documents, and then waiting for
the owner to approve the layout. This is a task that should have
been completed during the initial stages of the project. The project
engineer was responsible for requesting a further budget allocation of
hours for those groups who were involved in the development of the
standard format, which was not initially identified as a task for the
groups.
There were various forms of control that the project engineer
utilised on the project, the main form being the reallocation of per-
sonnel to suit the scheduled requirements. The other form of control
was to request from the owner, additional hours for work performed,
hours that were not included in the initial planning.
Monitoring
General
Monitoring forms an essential part of the (re)planning–implementing–
monitoring–reporting cycle. It encompasses the tasks of collecting and
recording information concerning the behaviour or performance of the
project; this will reflect how the project is going, at any given time, in terms
of production (work completed, attainment of milestones, . . . ), resources
(money) used and so on. This information is used in comparison with the
planned performance to influence the selection of updated controls.
Changes in scope, that is changes in the extent or nature of the work
itself, are additional considerations commonly monitored. Changes in scope
also feed into the replanning.
The Pareto rule could be expected to apply. It may only be necessary to
monitor 20% of project performance at any one time in order to get an
indicative feel for how the project is going.
Replanning 187
Data on performance is collected at the lower project levels and summed
to give performance at higher project levels.
Frequency of monitoring
The emphasis in monitoring is on providing information on a timely basis.
The frequency of monitoring is determined by considerations such as:

Duration of the project

Value of the project

Current project work volume; how many activities are underway at
any time

How fast the project is progressing

Degree of smoothness or number of troubles or complications on the
project; degree of difficulty of implementation

The project stage or phase

Resource availability to undertake the monitoring

The importance of the current activities

The stakeholders’ expectations

Organisation or industry norms or culture

The level in the organisation using the information.
Data collection
Monitoring involves data collection. Data collection practices may vary
between industries, between organisations, between individuals, between
in-house projects and outsourced projects, between owner and contrac-
tor and between delivery methods. Usually, however, information on the
following is required.
For (re)planning purposes

Production
– Equipment, people (labour) and materials

Resource usage
– Arrivals, departures, duration, costs, number (quantity) and type
– Equipment and people (labour).
Both are related back to activities and events for progress determination
purposes.
Changes in scope are also monitored.
188 Conventional treatment from a systems perspective
For project management purposes

Contracts – performance

Project overheads

Quality

People behaviour.
Computers assist in the storage of the huge amounts of data that are col-
lected as well as in the reporting function (creation of legible reports of
performance in absolute terms, and reports comparing planned and actual
and showing slippage, overruns and underruns). Forms for the manual col-
lection of data are designed with the view to transferring the data to a
computer. Data consolidation takes place before reporting. Information on
resources is converted to a money unit for cost reporting and replanning
purposes.
Example – People (labour)
Data on labour is collected using standard forms such as Figure 6.3. In
association with the labour data, data on production is also recorded such
as by the use of forms illustrated in Figure 6.4. The person in charge of the
workers is best placed to complete such forms.
Information from such forms can be used to establish production and can
be used to assist in estimating remaining durations to complete activities
currently in progress. The latter use facilitates replanning.
Data is regularly collated and developed into status reports. Labour costs
are related to the production.
Project
Date
Occupation Employee Activity
Code
Regular Overtime
Name No. Hours Pay
rate
Hours
Amount
Pay
rate
Figure 6.3 Form for collecting data on labour – daily ‘time’ sheet.
Replanning 189
Project
Date
Activity
Work item Measurement
unit
Planned Actual Latest revised
Figure 6.4 Form for collecting activity information.
Case example – Road project
In a road project, the first stage involved a bulk excavation activ-
ity of cut to fill. The operation used hourly hire plant from several
companies. The spread of equipment involved was:

Group 1: Scrapers, dozer, compactors, pad foot and flat drum
rollers and graders.

Group 2: Articulated dump trucks, excavator, graders, dozer and
compaction equipment.
Costs on the project compared with the budget were usually reviewed
on a monthly basis. The inherent variability of equipment posed a
problem with uncertainty in productivity. Activities, using methods
which could be susceptible to variable production in different ground
and stockpile conditions, required tighter monitoring of costs. Costs
were evaluated on a daily basis. If the production for the day was
different to the planned production, methods could be altered. It was
important to study daily cost data and format the data into production
costs, to increase confidence in planning and forecasting of activities,
for example a change over from articulated trucks and excavator to
scrapers and dozer.
On any one day the two groups of equipment could be working
on several different types of activities, so an overall measurement of
productivity would prove to be inaccurate. As a result, a practice
of docket submission by all operators of cartage plant (scrapers and
trucks), including load counts, and also recording labour used, mate-
rial moved and supervision, was implemented in order that the total
overall cost could be assessed.
190 Conventional treatment from a systems perspective
This method gave an estimate of the amount of material moved,
which was reconciled against the actual survey of the volume moved,
to give a close on the results.
0
1
2
3
4
1 3 4 5 6 7 8 9 10 11
Days
Actual costs
Budget
A
v
e
r
a
g
e

c
o
s
t

p
e
r

m
3
2
Figure 6.5 Example summary information.
An analysis of project production rates for the various cut to fill
operations was undertaken, for both overall cut and fill, and for
each spread of equipment. This last information could determine the
efficiency of certain operations and detected whether the excavators
and trucks were better suited to hauling from the stockpile or the cut,
compared with scrapers.
Cost codes were used to break down the dockets and the whole of
the information was entered on a spreadsheet to quickly analyse and
determine project production rates.
The information was presented in a manner that could be easily read
and understood by the project manager, engineers, superintendent
and foreman. This enabled all those directly involved, to understand
the performance of the work and enabled change or corrective action
to be taken. Figure 6.5 illustrates the type of summary information.
Popular approaches to reporting
Introduction
Reporting project performance follows the monitoring. The frequency of
reporting commonly matches the frequency of monitoring.
Reporting on performance at all project levels occurs. Each has its place.
Much reporting practice has been built up over many years, and is naturally
Replanning 191
ignorant of any systems approach such as advanced in this book. However,
all popular reporting can be related back to systems thinking.
Reporting may be to several levels of an organisation, possibly in corre-
spondence with the work breakdown levels, and the information is tailored
to the particular organisation level and its interests in the project. Thus,
reports will differ in frequency and depth. Not all reports are used for
planning; some are for information purposes only. Planners and managers
close to the workface need detailed information about individual activi-
ties. Reports at this level are frequent. Senior management would require
overview reports less frequently.
Reporting is carried out frequently enough to allow replanning to occur
before the project can get too far ‘off the track’. Timeliness of information
is important. (See section ‘Frequency of monitoring’.) The frequency of
reporting also takes into account that there is an inevitable time lag between
the recording of data and the processing of this data by computer into a
form useable as a decision tool by the planner or project manager.
Reporting could also be expected to occur at project milestones.
Flow-on aspects to assist management
Reporting well done, besides providing a basis for the selection of replanning
controls, gives rise to many flow-on aspects that assist the overall planning
and project management. For example:

It creates an awareness of the progress of concurrent activities, the
relationship between activities, and of coordination troubles.

It can provide early warning signals of potential troubles and delays.

Unacceptable work progress and practices are detected early, allowing
early correction.

Important needs of the project can be distinguished.

Owners and others not directly involved with the project work are kept
informed of project progress; it provides knowledge of profit and loss
throughout a project.

Involvement from/by various stakeholders can be highlighted.

Projects that are seen to be performing well can provide intrinsic moti-
vation to the project team.

Reporting can be considered part of the quality assurance process.

Regular and thorough reporting can be regarded as a discipline.

It provides a check on the accuracy of the planning.

It has a use in resolving disputes that may arise.

The estimating and planning of future projects benefit from the
information.
192 Conventional treatment from a systems perspective
Common reporting troubles
Common reporting troubles seen in practice include:

No report

Reporting for reporting’s sake

Too much reporting

Too much detail, both in reports and in data collected; this creates an
unnecessary burden on workers and supervisors who collect data; also
it is easy to miss important information in reports that are too detailed

Poor correspondence between the plan and that reported; the reporting
ideally tracks information directly related to the project plan

Self-deluding report, rose-coloured picture painted

Falsified report; disguised status of project

Assuming report

Not directed at the end-user

Lack of timeliness

Presence of gobbledegook

Poorly designed report.
A typical report
Comprehensive project reports contain information on:

Resource usage, cumulative resource usage, resource usage variance.
This is also converted to a common unit of money (cost).

Production, cumulative production, production variance. This is also
converted to information on schedule.
Also useful (and containing information on resource usage and production)
are:

Exceptions

Milestones

Updated project cost, cost variance and cost to go

Updated project completion time, production/schedule variance and
duration remaining

Contingency/delay allowance usage and variance

Incidents

Reasons for possible ills or health, for example unexpected events that
occurred. Suggested possible remedies for any ills.
This is in addition to graphical displays of progress:

Variance plot

A bar chart or time-scaled network
Replanning 193

S curves

Cumulative production plots

Cumulative delay plots.
And more general managerial/business issues:

Any outcomes or spin-offs anticipated.
The reports may be for purposes other than replanning, for example, to
take forward data to future projects, and to inform non-planners.
Where as-planned performance and actual performance are given, the
difference assists the selection of replanning controls. Some of this infor-
mation may be displayed in tabular form, but the graphical form tends to
be more readily comprehended (‘a picture is worth a thousand words’).
Reports provide both the positive aspects of a project’s performance and
the negative aspects.
The term ‘status report’ may be used by some people.
A variance report indicates the difference between the as-planned and
actual either in a raw unit (days, money, . . . ) or as a percentage. This may
be at the activity or project levels.
A milestone report indicates, for each project milestone, the planned value
(days, money, . . . ) for milestones yet to be reached, and completed values
and variances for milestones reached.
An incident report lists any current or immediate-past problems, their
potential impact on the project, and thoughts as to what corrective action
may be used.
Exception reporting involves only highlighting those occurrences that are
regarded as lying outside usual behaviour.
Cost reporting
Resource usage may be converted to a unit of money (cost), and cost
reporting may involve all or some of, or be involved in:

Variance reporting (budget minus actual)

Milestone reporting

Incident reporting

Exception reporting

Cost so far and cost to go.
A cost management summary something like Figure 6.6 might be used. In
order to satisfy a project’s cost considerations, management needs to be
aware of the money spent on a project compared with the spending planned.
As invoices and progress billings are received, commitments are trans-
ferred to incurred costs.
194 Conventional treatment from a systems perspective
Activity Committed Variations Total
(Forecast
cost at
completion)
Cost
estimate
Remaining
Total
Figure 6.6 Cost management summary.
S curves
S curves may be drawn for resource types (or money) as-planned and actual
(Figure 6.7). Acomparison of the two plots provides one means of reviewing
the performance of a resource type (or money).
There is a need to relate the information in Figure 6.7 with informa-
tion on production/schedule performance. For example, if the vertical axis
was cumulative money/expenditure, Figure 6.7 shows spending less than
planned. However, the diagram tells nothing of whether the project is ahead
of or behind production/schedule. Ideally, Figure 6.7 should be related to a
bar chart or time-scaled network showing progress to date. Alternatively,
state space reporting (Figures 1.6, 1.7 and 1.8), or the idea of earned value
may be used to relate expenditure and production/schedule.
Reporting resource (or money) performance on an S curve gives informa-
tion only at the project level, not the activity or lower levels. Original data
has to be examined to establish what is happening at the lower levels.
For the money case, the S curve shows cumulative money versus time, and
may be called cumulative project expenditure. Where there is also income
involved, for example a contractor being reimbursed for work done, the
Replanning 195
Project
start
Project
completion
Time
100%
0%
Actual
As-planned
(baseline)
Reporting
date
Variance
Underrun/underuse
(Overrun/overuse)
C
u
m
u
l
a
t
i
v
e

r
e
s
o
u
r
c
e
/
m
o
n
e
y
Figure 6.7 S curve or cumulative resource (or money) plot; as-planned and actual.
cumulative income plot can be superimposed on the cumulative expenditure
plot to give something like Figure 3.17.
Milestone reporting
Milestones are shown on a network or bar chart in order that everyone is
aware when particular important events occur, or when production should
be accomplished by. Replanning may be necessary should the milestones
occur late, or it becomes apparent that a milestone would not be reached
by the desired date.
Milestones may be signed off as they are completed in a summary of
milestones that can be consulted to determine the progress of the project.
Milestone summaries may be used by people managing several projects in
order to help minimise the degree of interference between projects.
Milestone reporting may be on resource (money) usage or on produc-
tion/schedule.
One criticism of focusing on milestones as a measure of assessing whether
a project is on track is that behaviour could be expected to follow Parkin-
son’s Law where the milestone date has been set too conservatively. Placing
greater emphasis on the final project completion date and lesser emphasis
on milestone dates plays down the influence of Parkinson’s Law. These
comments relate to the case where penalties are attached to not reach-
ing milestones as scheduled; where payment or incentives are attached to
reaching milestones, behaviour will be different.
Those accountable for both estimating and achieving milestone dates
quite naturally include some ‘time’ reserve in the estimate. Accountability
only for achieving a project completion date could be expected to lead to a
lesser total ‘time’ reserve in the project.
196 Conventional treatment from a systems perspective
One suggested practice, if milestones are to be used as a focus for report-
ing (and no incentives are in place), is to include the idea of a buffer as used
between activities in cumulative production plots and linear projects. Where
the milestone separates project stages, a buffer is incorporated between the
stages. This buffer reflects the conventional ‘time’ reserve in the milestone
date estimate. Use of this buffer is then reported in the same way as delays.
However, there is a view that tracking actual project progress against a
minimum estimated duration and a buffer may increase reporting for little
benefit.
Exception reporting
Exception reporting is analogous to practices used in statistical quality
‘control’. There, readings lying outside upper and lower tolerance limits
are regarded as matters for concern; readings lying within these limits are
regarded as due to natural variability and not acted upon. See for example,
the chart in Figure 6.8. In projects, only that information that is regarded
as an outlier is acted upon.
Tolerance ideas might be transferred to S curves, such that an enve-
lope of resource (or money) usage establishes the upper and lower limits
(Figure 6.9). These limits may be, for example, ±10%. The project is
allowed to proceed without adjustment as long as the actual performance
lies within the limits.
Schedule reporting
What is termed ‘schedule reporting’ is reporting on production. The report
of actual production/schedule performance is commonly presented as a tab-
ular status report (for example, Figure 6.10 gives two possibilities). Report-
ing may be selective and only occur on critical or near critical activities.
0 5 10 15 20
150
140
130
120
Sample number
Upper tolerance limit
Lower tolerance limit
Nominal dimension
M
e
a
s
u
r
e
m
e
n
t
s

(
m
m
)
Figure 6.8 Example statistical quality ‘control’ chart.
Replanning 197
Project
start
Project
completion
Time
100%
0%
Reporting
date
Lower limit Upper limit
Actual
As-planned
C
u
m
u
l
a
t
i
v
e

r
e
s
o
u
r
c
e
/
m
o
n
e
y
Figure 6.9 Limit envelope on an S curve.
In Figure 6.10b, the ‘Comments’ column records reasons for particular
entries, anticipated future delays and so on. All activities – physical, admin-
istrative, delivery etc. – are included. This may mean more than one person
completing the status report for any project.
The interpretation of the per cent complete column is important because
some items, though technically or physically not complete, can still allow
following activities to start. For example in building a house, roof tiles might
still require grouting, yet at the stage where all tiling except the grouting
has taken place this provides an adequate wet weather cover for the internal
building trades work to start. Of course, this fact could have been taken
into account in the scheduling by using an overlapping relationship between
the activities of roof tiling and internal trades work.
The information in the status report can be transferred to the network.
A time-scaled network or connected bar chart is more convenient for the
recording of the progress of the project. Some colour coding or shad-
ing scheme is useful for distinguishing between the as-planned and actual
performance.
Humour has it that: The more you fall behind, the more time you have
to catch up.
Bar charts
Bar charts are commonly upgraded by drawing, parallel to an activity
as-planned bar, a bar of length proportional to the amount of the activity
production completed, commencing at the date when the activity actually
started. Figure 6.11 shows two different drafting styles. A vertical line on
(a)
Project:
Date:
Estim.
Dur'n
Scheduled
%
Comp.
Rem.
Dur'n Activity Earliest
Start
Latest
Finish
Total
Float
Comments
Activities
in Progress Activity Comments
Start
date
Comp.
date
Start
date
Days in
Progress
Estimate of
days to Complete
(b)
Completed
Activities
STATUS REPORT
PROJECT ......................................................................
DAY ...............................................................................
Figure 6.10 Example activity status sheet/progress report/form.
Replanning 199
Time contour/timeline
(at reporting
date)
Activity ahead of
production/schedule
Activity behind
production/schedule
As-planned
Actual (as-executed)
(a)
AA
BB
CC
DD
EE
FF
Reporting
date (timeline)
As-planned
Actual (as-executed)
CC behind production/schedule
EE ahead of production/schedule
Remainder of activities on production/schedule
(b)
Figure 6.11 Bar chart reporting of progress; bar chart with status information.
a chart at any date indicates where the project and individual activities or
production should be up to at that date.
Replanning (in an iterative analysis sense) will constitute at least a new
network analysis leading to a new bar chart; the analysis is based on the
status report, using the current status as the initial conditions for the analysis
of the remainder of the project. However, if the intention is to stay with
the original program then this new analysis is not necessary.
Time-scaled networks are treated similarly.
200 Conventional treatment from a systems perspective
Reporting production/schedule performance on a bar chart gives infor-
mation only at the activity level, not the project level. Hence, conclusions
cannot be drawn from the bar chart alone as to how the project is perform-
ing with respect to production/schedule, though an overall ‘feel’ of how the
project is going relative to planned production or original schedule can be
obtained.
Factual networks
A factual network is also called an as-executed (as-built, as-constructed,
as-implemented) program. It is a network (best done on a timescale) of
as-executed activities indicating changes to the original program, including
delays and variations, and agreed changes, cross-referenced to files on this
information. It is a record of facts. The factual network becomes a per-
manent record of a project’s progress. A factual network (as opposed to
the original network) may not lend itself to analysis via conventional meth-
ods. The factual network also finds a use in attempting to resolve disputes
involving delays, time period extensions, liquidated damages and related
claims. It is similar in concept to an as-built drawing, in that it shows the
actual durations to perform the various work items (Figure 6.12).
Work changes, including variations, alterations, extras, omissions and
deductions, as well as delays can be argued on a more sound footing using
a network as a basis. If a network is included as part of the contract, then
all work changes and delays can be argued relative to this. Delays and
work changes affecting non-critical activities can be argued differently to
those affecting critical activities. Delays and work changes caused by the
(a)
(b)
Work
change Delay Delay
Work
change
Figure 6.12 Example network showing changes and delays. (a) Network as-planned;
(b) Network as-executed.
Replanning 201
owner or the contractor, or by both the owner and the contractor, or by
neither can be reconciled in principle far easier using a network of the
project as a basis than without such a network. Should also direct and
indirect cost structures form part of the contract, then the cost implications
of work changes and delays can be established more equitably for both
contract parties. Similarly time period extensions are placed on a more solid
footing.
The analysis of the effects of work changes or delays is done by com-
parison with the network which was the original plan of the project. The
factual network contains components that are the work activities as carried
out, together with components due to changes and delays. This enables
the responsibilities to be better ascertained for critical delays (delays affect-
ing the overall project progress). Such calculations are often carried out
manually although the computer coding of such an analysis is not a diffi-
cult task.
The factual network is supported with project records; the form adopted
for recording such information can be flexible but would include space for
cause, description, duration, other activities affected and comments.
Earned value
If a project is running ahead of production/schedule, then it would be
expected that the actual cost would exceed the planned cost at that point in
the project. Similarly, if a project is running behind production/schedule, it
would be expected that the actual cost would be less than the planned cost
at that point in the project.
Reporting project performance only on a conventional S curve (money)
may give a misleading impression of the health of a project because it
addresses only cost and not production/schedule. Earned value reporting
attempts to address this; it is a reporting device at the project level where
project performance with respect to both cost and duration is integrated.
The essence of earned value reporting is given here because of its use in
industry. However, reporting in a state space is a preferred way of reporting
(Figures 1.6, 1.7 and 1.8).
The earned value of production performed is found by multiplying the
estimated per cent completion for the production by the planned cost for
that production.
Earned value =(Estimated) Per cent complete ×Planned cost
This is the amount that should have been spent on the production at that
point in the project.
202 Conventional treatment from a systems perspective
Per cent complete and work packages
Estimating the per cent complete can be difficult to do with any precision,
unless a large amount of effort is dedicated to the job. It can be approached
approximately in a number of ways either at the activity level or by breaking
the project up into small work packages, and then by following optional
rules. Some example rules are:

A 50–50 rule, where the activity/package is regarded as 50% com-
plete if it has begun, and the remaining 50% is allocated when the
activity/package is fully complete.

A 0–100 rule, where the activity/package is only regarded as com-
plete if it is finished. (This rule will understate the per cent complete
because at any time there will be activities/work packages which are in
progress.)

Estimate per cent complete by ‘eye’ or judgement, depending on the
nature of the work, and the usefulness to following dependent activi-
ties/packages of stating percentage completeness of anything less than
100%.

Commercial computer packages may have their own peculiar rules.
Estimates made on work in progress will necessarily not be accurate.
Work packages in progress will have part committed costs and part esti-
mated/forecast costs. The total project cost will be a combination of esti-
mates, commitments and actual costs.
Complete activities/work packages may be referred to as ‘closed’;
in-progress as ‘open’; and unstarted as ‘unopened’.
When an activity or work package is complete, its original budget may
be referred to as being ‘earned’.
The selection of the appropriate duration of a work package is unclear.
The actual duration will relate to the overall project duration, the frequency
of reporting and the individual planner’s preferences. There is no definitive
answer as to what duration to choose. Shorter duration work packages will
provide more definiteness in estimating per cent complete because a greater
proportion of packages will be either complete or not started; there will be
a lower proportion in progress.
The plot shown in Figure 6.13 (an adaptation of S curve ideas) can be
drawn and provides a basis for evaluating cost and production/schedule
performance. The plot can alert management if variances are occurring that
deserve further attention.
In Figure 6.13, three plots are shown – actual cost, planned cost and
earned value. Reporting may give all three plots on one diagram as in
Figure 6.13.
Replanning 203
Value
completed
Actual
cost
Time
Cost
variance
Cumulative
cost
Production/schedule
variance
Planned cost
(baseline) Reporting
date
Figure 6.13 Example actual, planned and earned value plots.
Several types of variances can be identified in this figure. Two main
variances are:

Cost (spending) variance (underspent/overspent in a unit of money)
Cost variance, CV =Value completed−Actual cost

Production/schedule variance (ahead/behind production/schedule in a
unit of money)
Production/schedule variance, SV=Value completed−Planned cost
Note that the cost variance here is not the difference between planned and
actual. The values of the variances are only as accurate as the estimates of
per cent work complete.
Convention
The usual convention is for a variance to be negative if the project is behind
production/schedule or over budget (Table 6.1), but other conventions will
be found in the literature.
Table 6.1 Sign conventions on variances
Negative (−) Positive (+)
Cost variance Over budget Under budget
Production/schedule
variance
Behind
production/schedule
Ahead of
production/schedule
204 Conventional treatment from a systems perspective
Example
In Figure 6.13, at the reporting date, it has been estimated that the pro-
duction planned to be done by that date is 50% complete. That is, value
completed is 50% of planned cost.
This gives a negative production/schedule variance (indicating behind
production/schedule), and a negative cost variance (indicating over budget).
Supplementary reporting
Information on cost variances and production/schedule variances can be
put into more readily seen forms. Some examples are given here.
(1) Progress over the duration of a project might be plotted similar to
Figure 1.7, but with axes of cost variance (CV) and schedule vari-
ance (SV).
(2) Progress over the duration of a project might be plotted similar to
Figure 1.8, that is the change in cost variance and schedule variance
with time.
(3) Dimensionless indices or ratios may be defined and reporting done
in terms of these ratios or indices. The indices avoid the significance
of variances being influenced by the size of the project. For example,
a −$100, 000 variance on a $1 m project is more serious than on a
$100 m project.
Cost performance index (CPI) (CV ratio) =
Value completed
Actual cost
Production/schedule performance index (SPI) (SV ratio)
=
Value completed
Planned cost
The indices are particularly useful for establishing trends by plotting
their values as the project progresses (Figure 6.14).
In the CPI formula, planned cost and actual cost are being compared.
In the SPI formula, production performed and production scheduled
are being compared.
Indices or ratios of 1 or 100% indicate that the project’s progress
is on target. Indices greater than 1 indicate that the project is doing
better than expected. Indices less than 1 indicate unfavourable progress
(Table 6.2).
The SPI indicates production/schedule efficiency. For example, an SPI
of 0.75 or 75% indicates that only $75 worth of production has been
accomplished for each $100 of production scheduled.
Such indices or ratios also permit comparisons between projects.
Replanning 205
Date
1.0
1.5
0.5
0.0
x
x
x
x
x
SPI
x
x
x
x
CPI
Index or
ratio
value
Figure 6.14 Example; plots of cost and production/schedule performance indices/
ratios.
Table 6.2 Sign conventions on indices
Index < 1 Index = 1 Index > 1
CPI Over budget On budget Under budget
SPI Behind
production/schedule
On
production/schedule
Ahead of
production/schedule
Some earned value terminology not used in this book include BCWS
(budgeted cost of work/production scheduled; also called PV – planned
value), BCWP (budgeted cost of work/production performed; also called
EV – earned value), ACWP (actual cost of work/production per-
formed; also called AC – actual cost), STWP (scheduled time/duration
for work/production performed), ATWP (actual time/duration for
work/production performed) and cost/schedule ‘control’ system (C/SCS,
CSCS, CS
2
) because of their tendency to mislead. Earned value is sometimes
called performance, and the alternative name, performance measurement
(performance measuring technique, PMT), may be given by some people
to the earned value approach; this terminology is also not used here. Some
writers suggest, for reporting purposes, aligning work packages with project
deliverables; however, since there is a lot of confusion by writers and prac-
titioners over the use of the term ‘deliverable’, the term is best avoided.
Cumulative production plot
A cumulative production plot shows an activity’s or project’s production
up to any time or distance; the slope of the plot represents the rate of
production (Figure 6.15). Figure 6.15 can be either an activity level or
project level representation. A number of variants on Figure 6.15 can be
drawn (Figures 3.11, 3.12 and 3.13).
206 Conventional treatment from a systems perspective
Time
Time now
Actual Amount behind production
Learning
period
Planned rate
of progress/
production
C
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n
(
t
a
s
k
s
,

d
i
s
t
a
n
c
e
,

m
a
t
e
r
i
a
l

u
s
a
g
e
,

.

.

.
)
Time period behind
production
program
progress/
production
Figure 6.15 Example cumulative production plot.
Project
start
Project
completion
Time
Remaining
delay
allowance
Days
(buffer
available)
Rate of delay
allowance usage
Actual
delay
usage
Total delay allowance
Figure 6.16 Example cumulative delay plot.
Delay reporting
Some projects include a delay contingency (production/‘time’ reserve) and
monitor actual delays in production (Figure 6.16).
Case example – Land subdivision
This case example looks at one contractor’s approach to monitoring
and reporting on a land subdivision project.
Replanning 207
At the estimating/planning stage of the project the construc-
tion work is subdivided according to a work structure that identi-
fies/isolates similar work areas that are incorporated into the various
scheduled items. This enables a comparison at a later date with actual
costs. The selection of an appropriate degree of detail is considered
important when assigning cost codes to work areas. A more detailed
cost code breakdown would improve accuracy of the reporting but
may tend to make the whole process unwieldy and not user-friendly.
Daily timesheets are kept that record all plant, labour and materials
used for that day. These costs are proportioned to particular cost codes
that are allocated to the different items within the work breakdown
structure, for example earthworks, concrete work, and pipework bed,
lay and joint. From the information contained on these sheets it is pos-
sible to accurately calculate the cost of any particular itemand compare
the actual cost rates to the tendered price rates. These sheets can then be
totalled weekly and monthly to give the total costs for the month.
At the end of each month the costs that are established frominvoices
and dockets are totalled and printed out according to their allocated
cost code. With the use of a spreadsheet, the cost for the month is
compared to the budget for the month. From this, forecasts of budget
overruns and underruns can be made and where possible, action taken
to remedy any troubles.
Case example – Civil construction
Major items established at the commencement of the construction of
a wharf-ocean retaining structure project to assist with replanning
included:

Construction program

Cost centres based on construction program activities

Budget allocation for cost centres

Cash flow diagram for the project

Budget summary for cost centres

Personnel register

Internal/external plant registers

S curves

Activity production rates.
Construction program
The construction program showed the sequence of activities and dura-
tion of each activity, and assisted in the estimation of manpower and
208 Conventional treatment from a systems perspective
plant required at any stage, and was used to indicate whether the
project was behind or ahead of original schedule.
Cost centres
Major or summary activities, which separated the project into geo-
graphical areas or construction methods, were selected from the con-
struction program. A cost centre was then allocated to each activity,
and this gave a direct association between each activity and the budget
allocated to that activity.
Budget allocation
Once a cost centre had been allocated to each major construction
activity, resources were allocated to cost centres so as to estimate the
cost to complete each activity.
Account codes
Account codes chosen were:
10 Labour
20 Subcontractor
30 Permanent Material
40 Temporary Material
50 Internal Plant
55 External Plant
60 Fuel/Repairs
90 Overheads
Resources were allocated to account codes. This was converted to
costs per month and then multiplied by the activity durations to give
cost-to-complete. The durations were taken from the construction
program.
For recording of committed costs, purchase orders were filled out
for every dollar spent and a cost code allocated to that purchase order.
Details of purchase orders were entered into a costing program.
On receipt of invoice, delivery docket, purchase order and invoice
were matched, authorised for payment and sent to head office.
Cash flow diagram
The cash flow diagram was prepared at an early stage in the project
to assist the area manager in the management of cash flow for the
office.
Budget summary
The budget summary gave a lot of important information on one sheet
of paper because it indicated the project’s financial status such as
cost-to-date, cost-to-complete, budget and trends positive or negative.
Replanning 209
Each month, the cost variance was recorded. Any trends in this
variance were recorded and action taken, if necessary.
Personnel register
The personnel register provided the exact number of employees allo-
cated to each cost centre, and therefore was very useful when allocat-
ing labour costs.
Internal/external plant registers
The plant registers were similar to the personnel register in that they
gave information of exactly what plant was currently allocated to each
cost centre. At the beginning of each new reporting period, a purchase
order was written for every plant item and a cost code allocated,
giving a committed cost for the month.
S curves
For each cost centre of a construction activity, an S curve was estab-
lished to closely monitor progress. The plots had to be updated to
account for increases in quantities or approved time period extensions.
Activity production rates
Activity production rate information was used for pricing and the
programming of the project. Activity production rates were also used
for the pricing of variations during the construction period.
Replanning representation
Introduction
Planning uses open loop control incorporating estimates of future influ-
ences. This gives a project ‘baseline’ or ‘target’ performance. Disturbance
takes the project behaviour away from the current baseline.
As a precursor to replanning, monitoring and reporting of project per-
formance gives project status in terms of absolute performance and perfor-
mance relative to the baseline (errors or variances).
Replanning uses open loop control incorporating updated estimates of
future influences, and initial conditions equal to the current project state.
Error control ideas are used to guide the choice of which controls need to
be adjusted, but are not used in an automatic closed loop control sense in
current replanning practice. Replanning gives an updated project baseline or
target performance. Disturbance occurs. Monitoring and reporting follow,
and the cycle continues throughout the project.
Replanning is, in principle, no different to planning. Replanning controls
are no different to planning controls, and involve the selection of method,
210 Conventional treatment from a systems perspective
resources and resource production rates. However, note that changing the
work method may involve the development of a newnetwork and associated
flow on effects.
Summary
Project performance (the state x(li) is monitored on an ongoing (usually reg-
ular) basis, compared with planned performance and, if necessary, corrective
action is taken to send the project in a desired direction (on a desired state
trajectory). The controls bringing about corrections may be in the form
of method (including sequence), resource (or money) and/or resource pro-
duction rate inputs to the project. The controls are chosen to extremise
the objectives while satisfying any constraints present. Applying changed
or different controls amounts to replanning. Replanning comes up with
revised values of control variables, u(li, for the remainder of the project –
u(li has multiple parts (that is, it is a vector quantity of control variables)
containing information on method, resources and resource production rates.
Generally it can be said that prior work done constrains future (updated)
planning controls. Hence the project constraints are changing as a project
progresses. The flexibility in choice of controls decreases with time.
The following representation is given for the case where the project dura-
tion remains constant. A similar representation could be given where the
project duration is allowed to change.
Notation
The notation used in the following is:
N number of periods/stages
l period/stage counter, l =0, 1, 2, . . . , N–1
x(li planned-for state after period/stage l–1
x(0i initial state
x(Ni final/terminal state
xA(li actual state after period/stage l–1
x

(li revised planned-for state after period/stage l–1
var(li variance; difference between planned-for state and actual state
after period/stage l–1
u(li control at period/stage l
i(li disturbance; anything that prevents x(l+1i being obtained
exactly.
Baseline
The (planned) baseline is obtained by solving the deterministic optimal
control problem from project start to project end (or the equivalent itera-
tive analysis version). Future uncertain events/influences are converted into
Replanning 211
State/
output
x(k)
As-planned
u(0) u(1) u(N–1)
. . .
x(0) x(1) x(2) x(N–1) x(N)
ζ(0) ζ(1) ζ(N–1)
Period 0 Period 1 Period N–1
(a)
(b)
State space
State 2
State 1
As-planned
(baseline)
(trajectory)
0
1
2
N
x
x
x
x
x
x
x
x
Reporting dates
marked ‘x’
(c)
N–1
Stage 0 N–1 1
x(0)
x(1) x(1)
x(2)
x(N–1)
x(N)
Figure 6.17 (a) Project periods/stages. (b) (Planned) baseline for single state/output.
(c) (Planned) baseline for multiple (here 2) states/outputs.
equivalent deterministic influences. (There is no need to talk of risk manage-
ment.) The control is selected with the view that future system performance
is going to be affected by future events/influences. Figure 6.17a shows the
project broken up into periods. (The term ‘stage’ could also be used.) Each
period may be any chosen duration and can differ from other periods. At
212 Conventional treatment from a systems perspective
the end of each period, the project performance is monitored and reporting
is in terms of absolute performance, and actual performance relative to
planned performance. Figure 6.17b shows the (planned) baseline for the
case of a single state/output x(li, l = 0, 1, . . . , N. Such a representation
is only suitable for the single state/output case. For the case of multiple
states/outputs, a state space representation such as Figure 6.17c is necessary.
Figure 6.17c shows the (planned) baseline for the example situation of two
states/outputs, but it can be extended to higher dimensional state spaces for
multiple states/outputs.
Where there is a deliberate scope change part way through a project,
this will of course lead to a new baseline from the time of introduction of
the change onwards. The procedure for obtaining this new baseline is the
same as for the original baseline, except that the start time for the optimal
control problem is now the time of introduction of the changes.
Project performance better than planned, or as-planned
Should the project performance turn out to be better than planned, or
as-planned, then no replanning is necessary, though the planner may choose
to take advantage of any good performance, by replanning the remainder of
the project. As well, if new or better information comes to hand on future
influences, this can be used to refine the baseline for the remainder of the
project.
Project performance different to that planned
Replanning options are increased if a change in scope is allowed for the
remainder of the project. Allowing for a change in scope is equivalent
to allowing a change in the project (system) – a change in the box in
Figure 2.5. This in turn mirrors adaptive or learning system behaviour of
control systems theory.
Whether or not a change in scope is allowed, the original baseline can
serve as a target at every replanning point. Allowing a change in scope
enables adjustment of behaviour to match the baseline when needed. (If no
change in scope is allowed, the baseline (target) will change as the project
progresses.) Figure 6.18 shows the original (planned) baseline indicated
by x(li, l = 0, 1, . . . , N, and the actual behaviour indicated by xA(li, l =
1, 2, . . . , N. The actual behaviour xA(li will be different to the planned
behaviour x(li because of disturbance i(l–1i. The variance is the difference
between planned and actual,
var(li =x(li −xA(li l =1, 2, . . . , N
Replanning 213
State/
output
x(k)
var(1)
var(2)
Stage 0 1 N–1
(a)
x′(2)
State space
State 1
State 2 0
1
2
N – 1
N
x
x
x
x
x
x
x
x
Reporting dates
marked ‘x’
(b)
State 2
var(1)
State 1
var(1)
As-planned Replanned Actual
1A
x(0)
x(1) x(1)
xA(1) xA(1)
xA(2)
x(2)
x(N–1)
x(N)
Actual
Replanned
As originally planned
Figure 6.18 (a) (Planned and replanned) baseline and actual project performance
for single state/output. (b) (Planned and replanned) baseline and actual
project performance for multiple (here 2) states/outputs.
Or if no change in scope is possible, the actual behaviour xA(li will be
different to a revised (updated) planned behaviour x

(li because of distur-
bance i(l–1i. The variance is the difference between revised planned and
actual,
var(li =x

(li −xA(li l =1, 2, . . . , N
xA(li can be regarded as the initial conditions for (re)planning the remain-
der of the project. Commonly, the scope for this remaining part of the
project will be different to that envisaged in the initial planning. Solving
the deterministic optimal control problem from the current period to the
project end (or the equivalent iterative analysis version) will indicate what
scope has to be changed (if allowed) and what controls have to be adjusted.
214 Conventional treatment from a systems perspective
(If no change in scope is possible, this will give a revised baseline x

(li.) If
the original baseline is regarded as a constraint that has to be satisfied or
approached (or the final state – at the project end – is prescribed) in this
optimal control problem, the problem may be so heavily constrained that
little optimisation is possible. As before, uncertain future events are con-
verted into equivalent deterministic influences in order to do this calculation.
Clearly, solving such an inverse problem is difficult with present-day tools;
an iterative analysis-based mode of attack is necessary as the more pragmatic
approach using error control ideas to guide the selection of controls for the
iterations.
Allowing a change in scope would imply that the new scope may be
suboptimal with respect to the original project and end-product objec-
tives. However, the objectives and constraints may need to be modified in
the light of project information that has come to hand since the project
start.
Example
Consider a project where the planning outcome gave a state space plot as
in Figure 6.19. The line in Figure 6.19 is the baseline plot. The numbers
1, 2, . . . , 10 represent the 10 dates on which reporting is anticipated to be
carried out.
Performance of the project matches the as-planned performance up to
reporting date 3, but between reporting dates 3 and 4, actual performance
has differed from that planned because of various issues on the project. The
state at the latest reporting date is given by 4A in Figure 6.20. As-planned
and actual performance are given in Figure 6.20 along with the variances
in the states at the latest reporting date.
Cum
resource
usage
Cum
production
As-planned
(baseline)
1
2
3
4
5
6
7
8
9
10
x
x
x
x
x
x
x
x
x
x
Reporting dates
numbered
Figure 6.19 As-planned performance (baseline); example.
Replanning 215
x
x
Cum
resource
usage
Cum
production
1
2
3
4
5
6
7
8
9
10
x
x
x
x
x
x
x
x
x
Reporting dates
numbered
4A
As-planned
Actual
Variance in
cum resource usage
Variance in
cum production
Figure 6.20 Performance, actual and as-planned; example.
x
x
x
x
x
Cum
resource
usage
Cum
production
1
2
3
4
5
6
7
8
9
10
x
x
x
x
x
x
Reporting dates
numbered
4A
x
5A
Original baseline
Updated baseline
Actual
Figure 6.21 Updated baseline; example.
Replanning involves solving the planning problem with initial conditions
4A. The objectives and constraints may remain the same as for the plan-
ning problem or may have to be updated to reflect the latest thinking and
issues involving the project. Replanning gives the new baseline indicated
216 Conventional treatment from a systems perspective
in Figure 6.21; in this case, the replanning gets the project back onto the
original baseline by reporting date 6.
The final state of the project may change.
Exercises
1. How do you establish the best frequency of monitoring a project or
activities? Do you monitor at the completion of each activity? Do you
monitor after a certain time period? Other?
2. Review the information requested in any project forms to which you
have access. Comment on the design of the forms and the need for
more/less/alternative information.
3. List reporting problems, other than those mentioned here, of which you
are aware.
4. An owner, at pre-tender, estimates a contract package at $2.1 m. The
lowest bid from tenderers is $1.8 m and this bid is accepted by the owner.
Does the owner’s project manager administer the contract based on a budget
of $2.1 m or $1.8 m? Why?
5. It is tempting to compare actual costs with the budget. Consider
Figure E6.1, where the figure shows the actual cost running above the
budget at the latest reporting or status date.
What can be deduced from Figure 6.1 regarding:

How much is the project over- or under-spending compared with its
budget?

How much is the project ahead or behind production/schedule?
As-planned
Actual
Project
start
Project
completion
Reporting
date
Days
100%
0%
C
u
m
u
l
a
t
i
v
e

$
Figure E6.1 Example S curve.
Replanning 217
6. Give your views on the best way of obtaining a measure of the per
cent complete at any point in a project. You do not have to restrict your
thinking to the rules presented above.
7. In earned value reporting, the production/schedule variance is in a unit
of money. (This may seem strange.) Why do you think this is so? How
might you convert production/schedule variance to a unit of time?
8. What value is there for a cumulative delay plot that is additional to a
bar chart showing as-planned and actual? Or, is it redundant information?
Does it give management new information?
9. The terms ‘variance control’ and ‘configuration management’ are some-
times used to describe the process of watching the scope (of a project). Why
do some projects have a need to watch their scope? Do some contractors
deliberately play for variations?
10. Given that the early players in a project have the most ability to
influence the cost of a project, why is it then that examples continually
occur of projects proceeding without thorough preparatory work?
Is there a false economy here of project owners saving money initially in
the belief that the total project cost will consequently be less?
(This is reminiscent of the cartoon seen on notice boards in many engi-
neering offices where two ‘planners’ are looking at the drawings for the
Tower of Pisa in Italy [before it was built] and commenting that they could
save many Lire if they did not carry out soil tests. The building is now the
Leaning Tower of Pisa.)
11. What matters have to be taken into consideration in monitoring,
reporting and replanning with respect to cost? What do you regard as the
most important matter? What is commitment costing?
A
B
C
D
E
F
G
H
I
5
4
3
6
7
4
2
1
6
0
Figure E6.2 Example project network.
218 Conventional treatment from a systems perspective
Table E6.1 Example project data
STATUS REPORT
PROJECT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DAY End of day 10
Activity Completed
activities
Activities in
progress
Estimate of days
to complete
Comments
Duration
(days)
Days in
progress
A 6 Material supply
problem. Started
day 0
B 4 Started day 0
C 2 1 Started day 8
D 4 6 Bad weather delay
Started day 6
E 6 1 Started day 4
F 4 Started day 4
G 2 4 Started day 8
H (2) –
I (2) Revised estimate
12. The status report at the end of day 10 for a project (network shown
in Figure E6.2) is given in Table E6.1.
Draw a bar chart showing the original schedule and the activities as-
executed. Assume that the original duration estimates of the activities
remain unchanged, except where noted in Table E6.1.
Update the network calculations. Draw an updated bar chart for the
remainder of the project after day 10.
What remedial measures are possible to get the project back on produc-
tion/schedule?
13. For monitoring, reporting and replanning practices of which you are
aware, identify their strengths and weaknesses.
14. If a scope change occurs, is it correct to regard it as being incorporated
into the project? Or, does it create a new project (one with a re-defined
baseline)?
15. Comment on the suitability or usefulness of the following contractor’s
report, as a feed into replanning. Note that a contractor’s project report is
more selective than an owner’s project report.
Project . . . . . . . . . . . . . . . . . . . . . . . . Date . . . . . . ¡. . . . . . ¡. . . . . .
Project revenue
Original contract value $ . . . . . .
Plus variations Approved $ . . . . . .
Pending $ . . . . . .
To be submitted $ . . . . . .
Rise and fall – projected $ . . . . . .
Projected revenue $ . . . . . .
Margin calculation
Total project revenue $ . . . . . .
Less project cost $ . . . . . .
Contractor’s share of savings $ . . . . . .
Projected margin $ . . . . . .
Forecast final margin
Best case $ . . . . . .
Worst case $ . . . . . .
Margin trend
Original $ . . . . . .
Previous two reports $ . . . . . . $ . . . . . .
This report $ . . . . . .
Cash trend (POS/NEG)
Previous two reports $ . . . . . . $ . . . . . .
This report $ . . . . . .
Program status
Original completion date . . . . . . ¡. . . . . . ¡. . . . . .
Approved time period extension (days) . . . . . . . . . . . . . . . days
Adjusted completion date . . . . . . ¡. . . . . . ¡. . . . . .
Current deviation (ahead/behind) adjusted
date . . . . . . . . . . . . . . . days
Submitted and likely but unapproved
extensions of time period . . . . . . . . . . . . . . . days
Extensions of time period not yet submitted . . . . . . . . . . . . . . . days
Target completion date . . . . . . ¡. . . . . . ¡. . . . . .
Turnover
Turnover achieved this month $ . . . . . .
Cumulative turnover to date $ . . . . . .
Forecast turnover to end of project
Next month $ . . . . . .
Month 2 $ . . . . . .
Month 3 etc. $ . . . . . .
Total $ . . . . . .
Chapter 7
Selected topics
People issues
Introduction
The development in this book largely centres on the mechanics of planning.
The place of people in planning has been deliberately excised from this, in
order to present a clarity of approach. But of course people are involved in
planning, and whenever people get involved in anything, that thing takes
on inexplicable complications; what in principle may be straightforward
takes on new dimensions of complexity.
Human resource or people management is a very imprecise art, with
most literature on the topic being subjective, opinion and anecdote
based; it lacks objectivity and an ability to justify through the scientific
method. It also tends to be faddish because of this. The following notes
are no different in that alternative views can be expressed with equal
enthusiasm.
Safety, industrial relations and quality involve people issues, in partic-
ular, that influence planning. An organisation’s policies (general guides
for decision making and individual actions), procedures (detailed methods
for carrying out policies) and standards (degree of individual or group
performance defined as adequate or acceptable) also influence planning
practices.
Planning is part of a project manager’s responsibilities, but involves
all personnel in a project to some degree. In some cases, the responsi-
bility for preparing the main project plan is that of a program manager
assisted by a program team, under the overall direction of the project
manager.
There is a project role involving coordination and the interrelationships
of activities and project participants, and in providing continuity of work
for the project participants. When scheduling people as a resource, this may
be done only as a group, team or gang, rather than as individuals.
Planning well done and realistic can be motivational to the project team.
Selected topics 221
Involvement of people
Planning for projects ideally involves contributions from everyone who
will be involved in the implementation of the plan, although there may be
organisational difficulties with achieving this. In practice, there are people
who do the planning (the planners) and there are people – end-users or
doers – who implement the plans, yet it is believed that there is merit in
the latter group contributing their expertise to the planning and in being
consulted in the development of any plans.
It is sometimes said that the success of a plan is directly proportional
to the participation of those responsible for carrying out the planned
work. Commitment to the plan comes from an involvement in develop-
ing the plan, and gaining some ownership of the plan. There is even a
case in some circumstances of having decentralised planning; that is, the
people responsible for the performance of the project do the planning.
Generally, a plan should not be imposed on others; agreed realistic plans
are more likely to be successful. There can be a view of doers, unin-
vited to participate in the planning, that the plan represents achieving the
impossible.
The involvement of end-users in the planning can have additional pro-
ductivity and quality-related benefits such as:

Giving to them a better understanding of their job

Giving an indication of where they fit into the overall scheme of the
project

They gain confidence in the presumably more efficient and more effec-
tive work methods; they can assume that they are working smarter not
harder

The reasons for proper reporting, feeding into replanning, become more
readily understood

Reporting is seen to be for the project’s benefit rather than ‘big brother’
checking up on employee performance.
However, the planning process can be tainted from involving the end-users
in the planning because of their lack of appreciation of planning and the
concentration on their own details at the expense of the bigger picture.
There have also been many successful projects where the end-users had
no involvement in the planning. This leads to the question of how much
involvement of the end-users should there be in planning?
Where people have to accept the work of planners, without having an
involvement, the planners have to be realistic and credible, and take into
account the views of others. Planners may be part of a larger service depart-
ment within an organisation (Figure 7.1). This may create a them-and-us
attitude unless the planners are seen as genuine members of the project
team.
222 Conventional treatment from a systems perspective
Planning
department
Possible
barrier
Figure 7.1 Possible barrier between the ‘planners’ and the ‘end-users’.
Project personnel should be aware of planning principles and good plan-
ning practice, though they may only exercise a contributing role and may
not be directly involved in the mechanics of planning. Planning is a disci-
pline that leads to informed management.
As well, top or senior management commitment and support for a project,
and the awareness of this support by project personnel, is believed to be
a necessary ingredient of a successful project. It follows that senior man-
agement must provide the necessary resources in order to undertake the
planning.
The above refers to the structured planning carried out on projects. But
everyone on a project will carry out some modest, possibly unstructured,
planning of their own.
Work breakdown and initial planning
In initial planning, there may be a need to bring together various parties in
order to provide their specialist input, or to provide alternative views.
In developing a work breakdown there is a general feeling that people are
only familiar with one or two levels below the level of their position; beyond
that it is too remote and they need assistance from those contributing
at lower levels. (There are some people, though, who have risen through
organisational levels and who may be capable of going further.) Input from
people who are responsible for the work (and later monitoring the work)
may be necessary. This aligns the requirements of the end-users and the
planners.
Good practice sees that the work breakdown structure (WBS) is reviewed
by all people with an involvement in the work, and revised as necessary,
ensuring that any revisions are also reviewed.
Selected topics 223
Barriers to effective planning
Many of the barriers to effective planning are people related and people
attitude related. Improvements in planning can be gained by recognising
the existence of these barriers. Some barriers are:

Many people find thorough planning tedious, and would prefer to start
‘doing’ before giving planning enough consideration.

Many people feel confident relying on their experience and judgement
to handle troubles when they arise (‘crisis management’, ‘shooting from
the hip’), rather than working through planning in a thorough way
and making prior allowances for possible eventualities. People falsely
believe that thorough planning stifles initiative.

Many people resent the time extent and cost involved in thorough
planning, and do not connect project success with thorough planning.
Other work pressures and poor personal ‘time’ management skills mean
that little effort is devoted to planning.

Planning involves creativity and the generation of alternatives and ideas.
Many people do not have such skills. People constrain themselves to
doing something today the way it was done yesterday. Lateral thinking
is not undertaken, and in some industries is discouraged.

People have trouble converting the owner’s value system into work-
able objectives and constraints. The development of project objectives
and constraints is commonly poorly done. And everything in planning
depends on what the objectives and constraints are.

Senior management often has poor attitudes to thorough planning and
its cost effectiveness. This is reflected in poor support/resourcing of the
planning activity.
Responsibility matrices
For each activity, there are attendant people resource requirements. People
resource and organisational (including coordination) requirements, respon-
sibilities and authorities can be represented in a responsibility matrix/chart
form which may have rows corresponding to the activities, columns corre-
sponding to people resources and entries corresponding to responsibilities,
or similar (Figures 7.2, 7.3 and 7.4).
The matrices are also a useful way of communicating these responsibil-
ities, roles and duties to people involved in the project. Such charts show
where different project personnel interface and where managerial coordi-
nation is required. They relate activities to people, to groups of people, to
departments and so on.
It is important to note that everyone knows who is doing what, that
is, that they know the division of responsibility. Without a clear idea of
224 Conventional treatment from a systems perspective
Project
manager
Project
engineer
Planning
engineer
Office
engineers
Field
engineers
Subproject Activity
A1
A2
A3
A4
A5
A6
B1
B2
B3
B4
.
.
.
B
A
Responsibility
. . .
Legend for entries:
Responsibility ∆
Notification
Support s
Approval a
Consultation c
Supervision v
Figure 7.2 Example responsibility matrix/chart.
responsibility and activity assignment, people can become confused. People
double up on some jobs, while some jobs do not get done at all. Project
performance suffers.
Replanning issues
When replanning, besides the usual controls related to work method,
resources and resource production rates, there are a number of people-
related things that a project manager can try to improve a project’s perfor-
mance, such as:

Improved leadership and management

Improved coordination of workers and equipment.
Selected topics 225
Activity A
P
e
r
s
o
n

1
P
e
r
s
o
n

2
P
e
r
s
o
n

3
P
e
r
s
o
n

4
P
e
r
s
o
n

5
P
e
r
s
o
n

6
e
t
c
.
Activity B
Activity C
Activity D
etc.
5 2 4
3 5
6
1 1
Key
1 Actual responsibility
2 General supervision
3 Must be consulted
4 May be consulted
5 Must be notified
6 Final approval
Figure 7.3 Example responsibility matrix/chart.
Initiator
Responsibility
Approval
Activity
A1
A2
A3
.
.
.
Coresponsibility
A4
Enter title or
person's name
Figure 7.4 Example responsibility allocation.
Associated with monitoring and reporting may be regular project review
meetings.
It can be useful to tie monitoring in with the organisational structure of
the project so that employees do not perceive the monitoring as something
external and imposed upon them, but as an essential part of their work.
226 Conventional treatment from a systems perspective
Data collection is carried out by project staff who generally may rather
be doing something more tangible. Enthusiasm for filling in forms is not
there and so the need for the data has to be insisted on. Feedback in a
form noticeable to those collecting the data assists by demonstrating the
usefulness of the data.
It is recommended that regular reporting be part of the statement of
duties of project personnel, and all personnel connected with the project be
tied into the reporting system.
Honesty in reporting the true status is important for effective replanning.
The human frailty of embellishing the truth, in order that it reflects better
on the person responsible, has to be avoided.
Learning curves
Gaining the benefits from repetitive work can only happen with people
involvement. Increased experience permits the improvement.
Case example – Computerised catalogues
Reporting and human behaviour are the focus of this case example.
These are discussed with respect to a start-up software company that
specialised in the development of computerised component catalogues.
The company produced computerised component catalogues for
discrete part manufacturers. The company had eight employees – two
managers and six programmers.
Because of the small number of employees, very few formal pro-
cedures were ever implemented. Work orders and reports were com-
municated verbally, with no written records. There were no formal
project plans and work progressed according to the needs of the day,
as defined by the managers. The result was numerous late deliveries
of computerised catalogues, and the lateness only became apparent
as the deadline approached. There was also no data from which to
learn lessons and plan for similar projects, and consequently mistakes
were often repeated, not only by different programmers but also by
the same programmers.
An example of this related to the building of the setup program to
install a computerised catalogue. This task is a standard one and only
requires minor modifications for each catalogue (for example, com-
pany file, file paths and names). For a particular hydraulic component
catalogue, a problem was found regarding the manner in which the
setup program determined which system files to install and which
not to install. The routine used by the catalogue install program was
causing some of the client’s other programs to cease to work after the
Selected topics 227
catalogue’s installation. This situation was obviously very embarrass-
ing to the company and a workaround was found and implemented
as quickly as possible.
The next catalogue that was completed, in this case a vacuum
component catalogue, was prepared using the older version of the
installation program template and hence was sent to the client without
the new workaround. This mistake was only realised when the client
complained that several programs were locking up after installation
of the catalogue. The failure to properly report and learn from this
problem made the company look very unprofessional to important
clients. It also caused much frustration to the programmers who took
pride in their work.
More recently, a weekly production meeting was instituted, mainly
for the purpose of reporting in both directions. A summary of the
latest meeting was kept and progress could be determined from this.
This did not replace the informal, verbal communications but rather
added a formal element that was sorely missing, as well as a historical
record. In this way, the company could manage schedule and quality
matters relating to the development of computerised catalogues.
The company programmers were all middle aged, and were sensitive
to the issue of supervision. In one weekly production meeting, when
presented with the idea of keeping formal timesheets, one employee
compared management to an authoritarian government. Monitoring
and reporting were looked upon with suspicion and the whole issue
had to be handled with care.
In spite of this attitude, a monthly timesheet was adopted for mon-
itoring the cost of the catalogue development projects. This was cho-
sen as the method because it was not very intrusive and the worker
maintained influence over this task. The issue of dishonesty was not
considered to be great because of the personalities of the employees
and because, as the number of employees was small, management was
in constant contact with everyone. As the company grows, however,
it is understood that a more rigid system will have to be implemented,
such as an automatic schedule tracking program residing on the net-
work that can ‘see’ what programs and projects are being run on all
the computers. How the employees respond to such a system is yet to
be seen. It is foreseeable that some employees may opt to leave rather
than work under such supervision.
The company situation with respect to reporting and the employees’
reaction to both reporting and monitoring is very sensitive. Unfor-
tunately, the need for proper reporting and replanning in software
projects is critical in order to achieve a successful project. This conflict
creates a difficult problem for the project manager.
228 Conventional treatment from a systems perspective
Exercises
1. Where do the monitoring and reporting personnel fit in the organisa-
tional structure of a project team?
Can the monitoring and reporting be done by the planners?
What communication is necessary between the monitoring and reporting
personnel and the planners in order that the project plan is continually kept
up to date?
2. Suggest ways of making monitoring, reporting and replanning an
accepted part of the work environment.
3. Planning calculations indicate which are the critical activities. Following
disturbance, replanning may indicate changed critical activities; some non-
critical activities may become critical, while some critical activities may
become non-critical. How do you stop people from overreacting to such
changes?
4. If a project person is trying to hide something that is his/her fault, how
do you get accurately reported information on a project’s performance,
especially with a large span of management? Should the people (namely the
planning staff) who are going to use the project data be the ones who collect
the data? This would have an added benefit of avoiding communication
errors. What role could the equivalent of an independent third party or
auditor have in data collection and reporting?
Monte Carlo simulation
Introduction
Ordinarily, planning is converted from a true probabilistic problem to a
deterministic problem for tractability reasons (Figure 7.5). However, it is
possible to acknowledge the probabilistic nature of projects.
Project activities generally have variability associated with them, leading
to uncertainty in the planning formulation. The variability in the estimate
Planning problem
simplifications
Conversion from
synthesis to
iterative analysis
Hierarchical
breakdown of
a project
Conversion from
probabilistic to
deterministic
Figure 7.5 Simplifications to make the planning problem more tractable.
Selected topics 229
of an activity’s duration may arise from the work itself or from the esti-
mator having insufficient data to characterise the work exactly. This has
led to the consideration of techniques such as PERT (program evaluation
and review technique), and Monte Carlo simulation in the planning of
projects (in an iterative analysis form). These notes examine Monte Carlo
simulation. Monte Carlo simulation offers a practical way to incorporate
variability. The planning of a project should ideally take into consideration
the variability in the durations of the project’s component activities and
the consequent varying resource demands, but usually this is not done in
practice.
Monte Carlo simulation is a technique for analysing nearly anything that
contains probabilities, and not just project networks.
PERT, in various forms, has been around since about the 1960s, with
seminal credit usually being given to navy research and development work.
PERT introduces uncertainty into activity duration estimates. Planners are
asked to estimate activity durations based on optimistic, pessimistic and
most likely considerations.
Calculations are performed on an expected activity duration together with
an associated measure of the variability of this duration. It is a first-order
method. It involves approximations.
The main information coming from a PERT-type analysis is the probabil-
ity of completing a project on a particular date. Consequently, attention has
focused on Monte Carlo simulation as a more general stochastic technique
to provide additional information such as:

The probability of completing a project at a given cost

The probability that an activity will become critical or not

How to increase the probability of project completion by a specific date
by increasing resource numbers to some activities in the network

Probabilistic cost diagrams or cash flow diagrams which help the deci-
sion maker.
Monte Carlo simulation in the analysis of networks appeared in the 1960s.
The technique involves repeatedly generating sets of random numbers,
transforming these in accordance with desired probability distributions of
the activity durations, and obtaining samples of the relevant network results.
The results are treated statistically with presentations in the form of his-
tograms, and methods of statistical estimation and inference are applied.
For these reasons, the Monte Carlo method is also a sampling technique,
and as such shares the same problems of sampling theory, namely the
results are also subject to sampling errors. Generally, therefore, Monte
230 Conventional treatment from a systems perspective
Carlo simulation which results from finite samples are ‘not exact’ (unless
the sample size is very large). Monte Carlo simulation provides a significant
analysis tool useful in (the iterative analysis form of) planning. Most people
agree that the only way to realistically incorporate variability in activity
durations for large complicated networks is to use the technique of Monte
Carlo simulation. Monte Carlo simulation has the ability to use overlapping
relations without creating subactivities.
Monte Carlo simulation, in effect, converts a difficult probabilistic anal-
ysis problem into many simpler deterministic analysis problems. The steps
involved for networks are:

Generating activity durations for each activity (via uniform random
numbers)

Carrying out a conventional network analysis (incorporating overlap-
ping relationships, if applicable)

Repeating the above two steps many times

Collecting relevant statistics on important items, such as project com-
pletion dates.
The first step is one of data generation, the second step is analysis, while
the fourth step is bookkeeping.
Monte Carlo simulation is suited ideally to computer use, where large
amounts of data can be handled with ease. It is not a technique for use by
hand or where a closed-form solution is required.
The term ‘variance’ in the following refers to the probabilistic/statistical
variance.
Activity duration generation
The duration of each activity is described by some probability distribution
(Figure 7.6). Sampling from each activity’s distribution is accomplished via
the use of uniform random numbers.
Uniform random numbers are numbers that have an equal probability of
occurrence. The associated probability distribution looks like Figure 7.7.
Figure 7.7, in fact, is drawn for the standard uniform random variable
which takes values only between 0 and 1, and is distributed uniformly
between these values.
Drawing the second parts of Figures 7.6 and 7.7 next to each other and
to the same scale on the vertical axes gives Figure 7.8.
Without performing any mathematics, it appears reasonable that
x
t
= I
−1
Y
(:
t
i t =1, 2, . . .
y y
f
Y
( y) F
Y
(x)
y
i
F
Y
( y
i
)

Figure 7.6 Probability density function (PDF) and cumulative distribution function
(CDF) for a typical activity’s duration.
v v
f
V
(v)
F
V
(v)
0
1.0
1.0
1.0
1.0
v
i
v
i
Figure 7.7 PDF and CDF for standard uniform variate \.
y v
1.0
1.0
F
Y
(y
i
) = v
i
F
V
(v)
F
Y
(y)
v
i
v
i
y
i

Figure 7.8 Transformation between \ and Y.
232 Conventional treatment from a systems perspective
That is, if (:
1
, :
2
, . . . , :
n
i is a set of values from V, the corresponding
values for x
t
, t =1, 2, . . . , n, are obtained through this last expression. This
is sometimes referred to as the inverse transform method.
Some distributions allow the inversion of the cumulative distribution
function in closed form. For example the exponential distribution, where,
I
Y
(xi =1−e
−0x
= :
then
e
−0x
=1−:
or
x =
−ln :
0
Generally, however, the inversion of cumulative distribution functions is
not possible in closed form, or it leads to unattractive computations. Other
means are then used to transform between the standard uniform variate V
and the activity duration Y.
For an Erlang distribution of type · with mean 1¡µ, the transformation
commonly used is:
x = −
ln
·
¸
t=1
:
t
·µ
For a normal (Gaussian) distribution with mean µ and variance u
2
, the
transformation commonly used is:
x = u

12
¸
t=1
:
t
−6


Note that to generate a single sample from an Erlang or a normal distribu-
tion, more than one standard uniform random number is required.
For discrete random variables and where probability density functions
(PDF) are discretised, the transformation is based on the shape of the prob-
ability mass function (PMF). For example, consider an activity duration
with the probability mass function as shown in Figure 7.9.
Selected topics 233
5 6 7 8 9 10 11
7/25
4/25
6/25
5/25
3/25
2/25
1/25
Activity duration
(days) y
p
Y

(y)
Figure 7.9 Example probability mass function (PMF).
There,
P|Y =5] =0.16
P|Y =6] =0.28
P|Y =7] =0.24
P|Y =8] =0.16
P|Y =9] =0.08
P|Y =10] =0.04
P|Y =11] =0.04
That is, 16% of the area under the probability mass function corresponds
to Y = 5, 28% to Y = 6 and so on. It follows that 16% of the random
numbers generated can be allocated to Y = 5, 28% to Y = 6 and so on.
For :
t
lying between 0 and 1, it is convenient to adopt a transformation as
outlined in Table 7.1.
Table 7.1 Selected relationship between v and ,
Value of v
i
generated Corresponding value of y
i
0 ≤ v
|
< 0.16 5
0.16 ≤ v
|
< 0.44 6
0.44 ≤ v
|
< 0.68 7
0.68 ≤ v
|
< 0.84 8
0.84 ≤ v
|
< 0.92 9
0.92 ≤ v
|
< 0.96 10
0.96 ≤ v
|
≤ 1.0 11
234 Conventional treatment from a systems perspective
This is a numerical equivalent to the closed-form transformations exam-
pled earlier for the exponential, Erlang and normal distributions. For any
discrete random variable, each distribution is handled in this way, and a
unique transformation set up.
In the simulation literature, alternative methods of transformation for
both continuous and discrete random variables are given.
Uniform random numbers
The above generation of activity durations depends on there being a supply
of uniform random numbers. How are these numbers obtained?
There are numerous mechanical ways by which uniform random numbers
can be obtained, for example rolling a dice, tossing a coin or spinning a
roulette wheel. For a dice, the numbers 1 to 6; for a coin, the numbers 1
(heads) and 0 (tails); for a roulette wheel, the numbers 0 to 37 (or 38) all
have an equal probability of occurrence.
Uniform random numbers are also available tabulated in appendices of
many books on statistics. Calculators, spreadsheet packages etc. have inbuilt
random number generation facilities as well.
Such means of obtaining uniform random numbers are not convenient,
however, if a computer-oriented approach to analysing networks with prob-
abilities is sought.
An alternative way of generating uniform random numbers is to use
what are called ‘pseudorandom number generators’ which are mathematical
expressions/numerical algorithms. They produce numbers which are dis-
tributed almost uniformly, and for most project purposes any small differ-
ence between genuinely uniform and almost uniform is of no consequence.
One example of a pseudorandom number generator is (for example, Ang
and Tang, 1984)
r
t+1
= (ar
t
+ci(mod ni
which uses modulo arithmetic. The right-hand side translates as the remain-
der after (ar
t
+ci is divided by n. Let
l
t
=Int

ar
t
+c
n

Then
r
t+1
= ar
t
+c −nl
t
In these expressions, r
t
, t =0, 1, 2, . . . , are random numbers; a, c and n are
non-negative integers chosen by the user (c may sometimes be chosen as 0).
Selected topics 235
These randomnumbers may be normalised by dividing by the modulus n,
:
t+1
=
r
t+1
n
t =0, 1, 2, . . .
and these constitute a set of random numbers between 0 and 1 with the
standard uniform probability distribution.
Numbers so generated are not strictly randombecause the whole sequence
can be duplicated in a deterministic fashion. The numbers are accordingly
called ‘pseudorandom’.
The algorithm starts by assuming a seed value, r
0
, calculating r
1
, then r
2
,
then r
3
and so on.
For example, let a =3, c =1, n=7 and r
0
=1. The calculations proceed
as follows:
t =0 l
0
=Int

3×1+1
7

=Int

4
7

=0
r
1
=3×1+1−7×0 =4
:
1
=4¡7 =0.571
t =1 l
1
=Int

3×4+1
7

=Int

13
7

=1
r
2
=3×4+1−7×1 =6
:
2
=6¡7 =0.857
and so on.
This leads to the numbers in Table 7.2.
Note that this provides a range of random numbers between 0 and 6,
the numbers repeat with a cycle length of 6 and not all numbers between
Table 7.2 Example generation of random numbers
i r
i+1
v
i+1
0 4 0.571
1 6 0.857
2 5 0.714
3 2 0.286
4 0 0.0
5 1 0.143
6 4 0.571
7 repeats
236 Conventional treatment from a systems perspective
0 and 6 are obtained. The cycle length is related to n and the standardised
random numbers are obtained by dividing by n.
For a useable set of random numbers, n has to be large to avoid the
possibility of the same numbers recycling. Many texts suggest using, for
example, values of n about 2
30
to 2
35
. The value of r
0
is usually chosen
between 0 and n.
For the case c =0, n should not be a multiple of l, and r
0
should not be
chosen as 0, else all random numbers will become 0.
Generally, values of :
t
of 0 or 1 should be excluded from the generated
set of random numbers.
Example
Figure 7.10 is an elementary network chosen to illustrate all the steps
in applying Monte Carlo simulation to a network analysis. An activity-
on-node diagram is used although the technique is equally applicable to
both network types. The technique is not restricted by network size or
complexity.
Each activity in the network has its own unique distribution describing
the variability in its duration. It is assumed here that the distributions are
of the Erlang type. Activity duration data are given in Table 7.3.
START
END
A
B
C
D
Figure 7.10 Example network.
Table 7.3 Example network data
Activity Duration distribution
Erlang shape parameter Mean (days)
A 5 5
B 5 7
C 5 8
D 5 3
Selected topics 237
Step 1
The process starts by generating sufficient random numbers in order to
sample from the activity distributions. Since the four activities are each
modelled by Erlang distributions with shape parameters of 5, then 4×5 =20
random numbers are required.
Using one particular pseudorandom number generator, for example, the
first 20 random numbers (standardised to between 0 and 1) turn out to be:
0.0006, 0.0011, 0.1855, 0.1536, 0.6107, 0.3812, 0.4812, 0.1803,
0.0647, 0.0995, 0.6912, 0.7502, 0.7533, 0.8062, 0.7061, 0.0040,
0.0672, 0.1422, 0.4166, 0.0822.
Step 2
These 20 random numbers translate into 4 samples of activity durations.
For example, for activity A
i
A
= −
5
5
ln(0.0006×0.0011×0.1855×0.1536×0.6107i
=18.28 days
By the same process the sampled durations for activities B, C and D are
respectively,
i
B
=11.84 days, i
C
=2.41 days, i
D
=8.13 days
Step 3
A conventional critical path analysis of the network is carried out using
the activity durations calculated in Step 2 (Table 7.4). (This analysis may
incorporate overlapping relationships if applicable.)
This information is collected and summarised with similar information
at the end of the simulation.
The simulation now repeats Steps 1, 2 and 3 many times.
Table 7.4 Results of critical path analysis on example network, first cycle
Activity Dur. (days) EST EFT LST LFT TF FF IF Critical
A 18.28 0.0 18.28 0.0 18.28 0 0 0 ∗
B 11.84 18.28 30.12 18.28 30.12 0 0 0 ∗
C 2.41 18.28 20.69 27.71 30.12 9.43 9.43 0
D 8.13 30.12 38.25 30.12 38.25 0 0 0 ∗
238 Conventional treatment from a systems perspective
Step 1
Step 1 is repeated. That is, a further 20 random numbers are generated.
Using the same pseudorandom number generator, these are:
0.3969, 0.7478, 0.7133, 0.1255, 0.1332, 0.2639, 0.4867, 0.2745,
0.6665, 0.3310, 0.6276, 0.6699, 0.3876, 0.5898, 0.0266, 0.4522,
0.6871, 0.6808, 0.5742, 0.7621.
Step 2
Step 2 is repeated giving
i
A
=5.64 days, i
B
=6.80 days, i
C
=13.21 days, i
D
=1.43 days
Step 3
The network is analysed using the latest sampled activity durations
(Table 7.5).
Note that the critical path has changed. The path now runs through
activity C rather than activity B. As a general comment for all networks,
the critical path(s) may change from one analysis to the next throughout
the simulation.
The process again returns to Step 1.
Final step
Steps 1, 2 and 3 are repeated many times. Each time Step 3 is carried out
information on activity start times, activity finish times and activity floats
is collected. The final step summarises this collected information. More
information implies greater reliance can be placed on the results and so it
is not uncommon for Steps 1, 2 and 3 to be repeated 1000, 10 000 or more
times. It is tedious work but work at which a computer excels. Computer
run durations are obviously longer than in the deterministic case but even
with large networks, the computer run duration is not a concern.
Table 7.5 Results of critical path analysis on example network, second cycle
Activity Dur. (days) EST EFT LST LFT TF FF IF Critical
A 5.64 0.0 5.64 0.0 5.64 0.0 0.0 0 ∗
B 6.80 5.64 12.44 12.05 18.85 6.41 6.41 0
C 13.21 5.64 18.85 5.64 18.85 0.0 0.0 0 ∗
D 1.43 18.85 20.28 18.85 20.28 0.0 0.0 0 ∗
Selected topics 239
Commonly the collected information is summarised in histogram form.
For example consider information collected on the project completion times
(the finish times of activity D). After o passes (o = 1000, 10 000, etc.)
through Steps 1, 2 and 3, there are o values of project completion times. By
usual means these values are converted to a histogram such as Figure 7.11.
Generally as o gets larger, it could be expected that the shape of the den-
sity function for the project completion time approaches more closely the
shape of the density function for the project completion time had the net-
work been analysed in a closed-form mathematical fashion. It is generally
acknowledged though that a closed-form analysis is difficult for all but the
most elementary networks with elementary assumptions on the probability
distributions describing the activity durations.
From Figure 7.11 it is possible to evaluate details such as the probability
that the project will be completed by a certain date. Probabilities of com-
pletion by a certain date are calculated from the area of the histogram to
the left of that date; probabilities of exceeding a certain date are calculated
from the area of the histogram to the right of that date.
For projects containing many activities, this histogram could be expected
to approach a normal distribution in shape according to the central limit
theorem. This is more so for a cost histogrambecause all activities contribute
to the total cost.
Criticality index
Monte Carlo simulation allows the calculation of a criticality index (Van
Slyke, 1963). The criticality index is the probability that an activity will be
on the critical path. The criticality index is defined as W(ai¡M, where
W(ai =
M
¸
t=1
u
t
(ai
Project completion time (days)
N
u
m
b
e
r

o
f

o
c
c
u
r
r
e
n
c
e
s
40 5
Figure 7.11 Example histogram of project completion times.
240 Conventional treatment from a systems perspective
and
u
t
(ai =
¸
1 if activity a is critical during simulation experiment t
0 otherwise
for M simulations.
For example, if an activity is on a critical path 3000 times out of 10 000
simulation runs, this activity is given a criticality index of 0.30.
Monte Carlo simulation can accommodate three options to establish the
probability of any activity becoming critical, that is the criticality index:
(i) The latest finish date of the project is set equal to its earliest finish date.
In this case, the non-critical activities in the network have positive
total floats and every network analysis contains critical activities.
(ii) A planned finish date is assumed and the latest finish date of the
project is set equal to that date. In this case, some activities may have
a negative total float and in some simulations the network contains
no critical activities.
(iii) This option is similar to the second option, the difference being that
the planned project finish date is not given but rather deduced from
a given number of simulations.
Option (i), in general, gives a higher criticality index. This follows from
the observation that a critical path occurs in every simulation. This is to be
compared with Options (ii) and (iii) where critical paths do not occur in
every simulation.
In Option (i), the Start and End activities have criticality indices of 100%;
that is, these activities have a probability of 1 of becoming critical. The Start
and End activities in Options (ii) and (iii) usually have probabilities less
than 1 of becoming critical; this value, however, is higher than probabilities
associated with other activities becoming critical.
Option (i) leads to positive total floats, while Options (ii) and (iii) may
give positive and negative total floats.
Cost inclusion
There are a number of ways by which costs can be included in a Monte
Carlo simulation. The approach given here is that described in Al-Sadek
and Carmichael (1992).
Each activity is assigned a mean value of cost, C
n
, and a ratio, B, of
labour and equipment cost to material cost. For example if B = 0.1, this
means that the cost of the activity mainly consists of material, while if
B =0.9 the cost of the activity mainly consists of labour and equipment.
Selected topics 241
It is further assumed that the cost of an activity is proportionally related
to the activity duration at a rate equal to B. This is based on the assumption
that the cost of materials stays constant but that the cost of labour and
materials varies as the activity duration varies. Also, it is assumed that
the resource numbers used for any activity do not change even though the
activity duration changes. That is,
C
a
= C
m
+
(T
a
−T
m
i
T
m
BC
m
where
C
a
is the cost of the activity at a given duration T
a
T
m
is the mean activity duration (Figure 7.12).
For each activity, a duration T
a
is generated fromits associated probability
density function. This also establishes C
a
from the above equation. Usual
forward and backward passes are then performed on the network with
earliest times, latest times, floats and costs being evaluated for each activity.
Example
Consider the construction of a three-span bridge with precast concrete
beams and piled foundations (based on Carmichael, 1989). Table 7.6 lists
the component activities, their mean estimated durations (T
m
i, their shape
parameters for the assumed probability (Erlang) distribution, the ratios of
labour and equipment to materials cost (Bi and the mean values of cost
(C
m
i. The time unit is day; costs are expressed in a suitable money unit.
The activity-on-node diagram for this project is given in Figure 7.13. The
translation of the activity codes used in Figure 7.13 is given in Table 7.6.
T
m
T
a
C
m
C
a
R = 1
R = 0.5
R = 0
Figure 7.12 Relationship between the cost and the duration of an activity.
Table 7.6 Three-span bridge data. (Activity names have been abbreviated because of
space limitations – work items are implied.)
Activity code Activity T
m
Erlang Shape
Parameter
R C
m
Establishment activities
Start Start up 1 10 0.8 1000
Detour Set up road detour 1 10 0.5 14 500
Site Prep. Site preparation 5 5 0.5 10 000
Order and delivery activities
O/d Piles Piles 40 20 0.2 61 000
O/d Rein. Reinforcement 5 20 0.2 11 000
O/d Forms Formwork and other
concrete materials
5 20 0.2 55 000
O/d Rail. Railings and guard rails 20 20 0.2 21 000
O/d Deck Precast deck beams 30 20 0.2 28 500
Earthworks
Excav. A Excavation for
abutment A
5 5 0.8 25 000
Excav. B Excavation for abutment B 5 5 0.8 25 000
Backf. A Backfill and compact
abutment A
4 5 0.8 32 000
Backf. B Backfill and compact
abutment B
4 5 0.8 32 000
Piling
Pile A Drive piles abutment A 8 10 0.8 12 000
Pile B Drive piles abutment B 8 10 0.8 12 000
Pile 1 Drive piles pier 1 7 10 0.8 12 000
Pile 2 Drive piles pier 2 7 10 0.8 12 000
Concrete, formwork and
reinforcement placing
Conc. A Abutment A 10 15 0.5 16 500
Conc. B Abutment B 10 15 0.5 16 500
Conc. 1 Pier 1 8 15 0.5 16 500
Conc. 2 Pier 2 8 15 0.5 16 500
Precast concrete placing
Deck 1 Deck from abutment A to
pier 1
4 20 0.85 8000
Deck 2 Deck from pier 1 to pier 2 4 20 0.85 8000
Deck 3 Deck from pier 2 to
abutment B
4 20 0.85 8000
Finishing activities
Surface Placing surface and fixing
railings and guard rails
10 15 0.8 18 500
Approaches Approach roads 5 10 0.8 5000
Slopes Slopes to approach roads 2 10 0.8 13 000
End Wind up 1 10 0.8 1000
Selected topics 243
O/d
Forms
O/d
Rail.
O/d
Deck
O/d
Piles
Pile
1
Pile
2
Start
Site
Prep.
Detour
O/d
Rein.
Conc.
1
Conc.
2
Excav.
A
Excav.
B
Pile
A
Pile
B
Deck
1
Deck
2
Deck
3
Surf-
ace
End
Slopes
Approa-
ches
Backf.
A
Backf.
B
Conc.
A
Conc.
B
Figure 7.13 Example three-span bridge project.
The simulations are described in Al-Sadek and Carmichael (1992). Sim-
ulation output constituted the following:

A histogram of the duration of the project. This is used to evaluate
the probability of completing the project by a certain date. Confidence
intervals may also be established from this histogram.

A histogram of the project cost. This is used to evaluate the probability
of completing the project on a certain budget. Confidence intervals may
also be established.

Criticality indices.
For 50 000 simulations, Figure 7.14 shows the histogram of the project
duration, and Figure 7.15 shows the histogram of the project cost.
Table 7.7 lists the criticality indices. The first column corresponds to
Option (i) where the project latest finish time equals the project earliest
finish time. The second and third column of values correspond to Option
(ii) with, respectively, a 90-day planned finish and an 80-day planned finish.
The fourth column of values corresponds to Option (iii) where the project
latest finish time equals the average value of the project duration (111.2
days by simulation).
244 Conventional treatment from a systems perspective
0
2 000
4 000
6 000
8 000
10 000
12 000
14 000
N
u
m
b
e
r

o
f

o
c
c
u
r
r
e
n
c
e
s
Duration (days)
0 100 200 300 400
Figure 7.14 Example histogram of the project duration.
250 500 750 1000
Cost (×1000)
N
u
m
b
e
r

o
f

o
c
c
u
r
r
e
n
c
e
s
0
2 000
4 000
6 000
8 000
10 000
12 000
14 000
Figure 7.15 Example histogram of project cost.
Variance reduction
The final step in Monte Carlo simulation involves collecting relevant statis-
tics on important items such as project completion times. This is basically a
sampling process, and sampling errors decrease as the sample size increases.
Monte Carlo simulation also permits the use of variance reduction tech-
niques which reduce errors or variances, and these can be used in place of
increasing the sample size from increased simulations.
To demonstrate this, consider the project completion time, T. For inde-
pendent sampling, the variance of the estimate
˜
T, for J samples, is
var|
˜
T] =
var(T
1
i +var(T
2
i +· · · +var(T
J
i
J
=
Jvar(Ti
J
2
=
var(Ti
J
Table 7.7 Criticality indices for example project
Activity Criticality indices (%)
Option (i) Option (ii) Option (iii)
80 day 90 day
Establishment activities
Start up 100.00 65.04 77.07 41.14
Set up road detour 0.03 11.64 18.91 4.35
Site preparation 2.32 13.34 21.79 4.91
Order and delivery activities
Piles 83.01 60.54 71.0 38.24
Reinforcement 0.18 14.29 21.87 5.10
Formwork and other concrete
materials
0.00 14.53 22.54 5.13
Railings and guard rails 1.47 2.01 3.97 0.69
Precast deck beams 12.99 10.09 15.30 4.17
Earthworks
Excavation for abutment A 2.35 13.41 21.91 4.92
Excavation for abutment B 0.99 7.46 12.57 2.26
Backfill and compact
abutment A
15.49 53.25 62.75 31.08
Backfill and compact
abutment B
85.54 60.73 71.60 38.27
Piling
Drive piles abutment A 30.07 58.53 69.81 36.72
Drive piles abutment B 7.83 54.54 67.68 33.71
Drive piles pier 1 83.01 60.54 71.00 38.24
Drive piles pier 2 28.07 59.15 70.49 37.25
Concrete, formwork and
reinforcement placing
Abutment A 77.71 60.28 70.80 37.64
Abutment B 70.05 58.74 70.82 37.56
Pier 1 54.51 59.27 70.02 36.02
Pier 2 64.36 59.64 70.31 36.63
Precast concrete placing
Deck from abutment A to
pier 1
13.56 54.49 68.25 31.43
Deck from pier 1 to pier 2 13.56 54.59 68.25 31.43
Deck from pier 2 to
abutment B
25.47 62.92 75.27 39.62
Finishing activities
Placing surface and fixing
railings and guard rails
98.48 64.99 77.02 41.14
Approach roads 73.11 60.15 71.37 37.70
Slopes to approach roads 1.52 49.53 61.76 29.25
Wind up 100.00 65.04 77.07 41.14
246 Conventional treatment from a systems perspective
That is, as J increases, the variance decreases and the estimate
˜
T gets
closer to the true value I(Ti. Doing calculations for large J may, however,
not be feasible and alternative methods may be invoked to improve the
efficiency of the computations. Consider one such method, namely that of
antithetic variables (for example, Graver and Thompson, 1973).
For every uniformly distributed random number :
t
, an additional random
number, 1−:
t
, is generated. Sets of random numbers :
t
, t =1, 2, . . . , give
rise to T as before. Sets of random numbers 1−:
t
, t = 1, 2, . . . , give rise
to T

in a similar fashion. T and T

are antithetic realisations.
The estimates become,
˜
T

=
1
J

T
1
+T

1
2
+
T
2
+T

2
2
+· · · +
T
J
+T

J
2

and
var(
˜
T

i =
1
2J
|var(Ti +cov(T, T

i]
The covariance term is negative because T and T

are negatively correlated
(antithetic). Large values of T will generally be associated with small values
of T

and vice versa, because of the underlying choice of the uniformrandom
numbers; if :
t
is large, then 1 −:
t
is small and vice versa. The negative
covariance term should lead to a smaller variance estimate for the project
completion time than by direct random sampling.
References
Al-Sadek, O. and Carmichael, D. G. (1992), On Simulation in Planning
Networks, Civil Engineering Systems, Vol. 9, pp. 59–68.
Ang, A-H. S. and Tang, W. H. (1984), Probability Concepts in Engineering
Planning and Design, Vol. II, New York: Wiley.
Graver, D. P. and Thompson, G. L. (1973), Programming and Probability
Models in Operations Research, Monterey: Brooks/Cole.
Otnes, R. K. and Enochson, L. (1978), Applied Time Series Analysis, New
York: Wiley.
Van Slyke, R. M. (1963), Monte Carlo Methods and the PERT Problem,
Operations Research, Vol. 11, No. 5, pp. 839–861.
Exercises
1. Consider a triangular distribution (probability density function) for an
activity’s duration as in Figure E7.1a where, i
1
is an optimistic duration,
i
3
is a pessimistic duration and i
2
is the most likely duration. Note, there
Selected topics 247
t t
f
T
(t) F
T
(t)
1.0
t
1
t
2
t
3
t
1
t
2
t
3
(a) (b)
2
t
3
– t
1
t
2
– t
1
t
3
– t
1
Figure E7.1 PDF and CDF for activity duration.
is a discontinuity at i = i
2
. Figure E7.1b shows the associated cumulative
distribution function.
Show that the relevant transformation between the standard uniform
variate V, and T is
i = |(i
2
−i
1
i(i
3
−i
1
i:]
1¡2
+i
1
0 - : ≤
i
2
−i
1
i
3
−i
1
i = i
3
−|(i
3
−i
2
i(i
3
−i
1
i(1−:i]
1¡2
i
2
−i
1
i
3
−i
1
- : -1
(Hint. You may find the integration of the line equation of the first part
of the probability density function for T easier if you first shift the origin
of the line to i = i
1
, and the integration of the line equation for the second
part of the probability density function for T easier if you next shift the
origin to i = i
3
.)
2. The following is a FORTRAN listing of a main program and subroutine
based on the pseudorandom number generator,
r
t+1
= ar
t
+c −nl
t
where r
0
=673, a =17, n =2
20
=1048576 and c =1.
PROGRAM SIM
. . .
CALL ATRAND(1,X,IR)
. . .
CALL ATRAND(0,X,IR)
. . .
END
248 Conventional treatment from a systems perspective
SUBROUTINE ATRAND(K,X,IR)
IF(K.EQ.0) GOTO 1
IR =673
1 CONTINUE
IA =17
IM =1048576
IC =1
IK =INT((IA∗IR+ICi¡IMi
IR =IA∗IR+IC−IM∗IK
X =FLOAT(IRi¡FLOAT(IMi
RETURN
END
Use FORTRAN or convert to a coding of your choice to generate, say,
1000 random numbers between 0 and 1.
Calculate the sample average,
¯ : =
1000
¸
t=1
:
t
and sample variance,
s
2
=
1
999
1000
¸
t=1
(:
t
− ¯ :i
2
of these 1 000 numbers. The sample average should be approximately 0.5.
The sample variance should be approximately 1/12.
Draw a histogram of these 1 000 numbers. The choice of interval size on
the horizontal axis is up to you. How close is the histogram in shape to a
rectangle?
3. The following is a FORTRAN listing of a main program and subroutine
based on the pseudorandomnumber generator given in Otnes and Enochson
(1978, p. 414).
PROGRAM SIM
. . .
CALL TDRAND(1,X,I)
. . .
CALL TDRAND(0,X,I)
. . .
END
SUBROUTINE TDRAND(K,X,I)
IF(K.EQ.0) GOTO 1
I =783637
Selected topics 249
1 CONTINUE
I =125∗I
I =I −(I¡2796203i ∗2796203
X =FLOAT(Ii¡2796202.
RETURN
END
Do calculations similar to the previous exercise.
Without using any involved statistical method, which pseudorandom
number generator, ATRAND or TDRAND, in your opinion gives the better
approximation to uniform random numbers?
4. Search the simulation literature for pseudorandom number generators
other than the one given above.
Linear projects
Introduction
Many projects are ‘linear’ in nature and repeat the same sequence of tasks
many times throughout the project lifetime. An example is the construction
of multistorey buildings where the same trades (for example, involving the
plumber, electrician, ceiling fixer and painter in sequence) are repeated floor
by floor. Other examples include pipeline and railway construction and
estate housing projects.
This linearity is exploited by cumulative production plot-based
approaches (activity level representation). The unit of production may be,
for example, kilometres, storeys or material usage. Such a plot shows an
activity’s production up to any time or distance; the slope of the plot rep-
resents the rate of production (Figure 7.16).
Learning
period
Planned rate
of progress/
production
Time or distance
C
u
m
u
l
a
t
i
v
e

p
r
o
d
u
c
t
i
o
n
(
t
a
s
k
s
,

d
i
s
t
a
n
c
e
,

m
a
t
e
r
i
a
l

u
s
a
g
e
,

.

.

.
)
Figure 7.16 Example cumulative production plot.
250 Conventional treatment from a systems perspective
Some variants on Figure 7.16, at the activity level, are:

The axis labelling may be interchanged, with the independent variable
on the vertical axis

The diagram may be drawn down the page

The line plot may be given a finite width where work occupies durations
of significance.
Cumulative production plots contain bar chart information integrated. It
is the additional vertical scale that makes the cumulative production plot
attractive, compared to a bar chart. The computational path to the cumu-
lative production plot should be no different to that of a conventional
bar chart, namely via a network. A network helps to clarify the logic and
dependencies between the activities.
Line of balance thinking applies to linear type projects; this is a process-
oriented approach which has its origins in the manufacturing industry.
In cumulative production plots, a resource schedule is one determined
by resource numbers (quantity), while a parallel schedule is one deter-
mined by a desired activity production rate. Activity production rate and
activity resource usage are directly related; activity production rate can be
increased/decreased by the use of more/less resource numbers.
Parallel scheduling involves lines (representing activities) being drawn
parallel to each other, with or without a buffer between them. To ensure
such activity production rates, the resource numbers are increased (or
decreased) accordingly. Such a scheduling practice might be adopted if there
is a need for a reduced project duration or a desired completion date. Paral-
lel scheduling will lead to shorter project durations but this is at the expense
of using additional resource numbers.
Resource scheduling is where the schedule is determined by the available
or usual resource numbers. At the planning stage, the start or finish times of
activities are adjusted in order to avoid interference between activities. Such
a scheduling practice might be adopted where the least cost solution is sought
irrespective of the project duration. Activities with the smaller line slopes can
be speededupby the employment of extraresource numbers andthis will have
the effect of reducing the overall project duration, but at an additional cost.
Separation or buffer time periods are selected to prevent interference
between adjacent activities, and are represented by gaps between the activ-
ities in a cumulative production plot.
Concurrent activities may be overlapped but this leads to a cumulative
production plot presentation that may be harder to read.
Learning
Repetition of work generally leads to later activities being done in shorter
durations than the same activities done earlier. Workers become more adept
Selected topics 251
at their jobs with experience. Such behaviour is included in a study of
learning and learning curves. Taking learning into account will bend the
lines in the cumulative production plot to being closer to vertical as the
number of repetitions grows.
Examples
Example
Consider two activities, in a project, namely ‘excavate trench’ and ‘lay pipe’.
Figure 7.17 shows the bar chart and associated time-chainage chart, where
the activities are proceeding at the same rate.
Should the second activity proceed at a faster rate than the first, the
situation in Figure 7.18 might occur.
To avoid the situation that occurs in Figure 7.18, namely that the program
says to lay the pipes before a trench has been excavated, a finish-to-finish
(F/F) relationship might be used, instead of a start-to-start (S/S) relationship,
as shown in Figure 7.19.
Example
Consider an earthmoving project involving cut, and installation of drainage.
The bar chart and time-chainage chart may look like Figure 7.20.
Example
The excavation of drains is sometimes carried out from the bottom of a
hill upwards, or roads are constructed simultaneously from each end. In
Excavate trench
Lay pipe
20 days
20 days
Excavate trench
Lay pipe C
h
a
i
n
a
g
e
0 3 20 23
Days
S/S lead time
period 3 days
Figure 7.17 Activities advancing at the same rate, and S/S relationship.
252 Conventional treatment from a systems perspective
Excavate trench
Lay pipe
20 days
Excavate trench
Lay pipe
0 3 20
Days
12 days
15
S/S lead time
period 3 days
C
h
a
i
n
a
g
e
Figure 7.18 Activities advancing at different rates, and S/S relationship.
Excavate trench
Lay pipe
20 days
Excavate trench
Lay pipe
C
h
a
i
n
a
g
e
0 20
Days
12 days
8
F/F lead time
period 0 days
Figure 7.19 Activities advancing at different rates, and F/F relationship.
such cases the time-chainage charts might look like Figure 7.21. Note the
orientation of the chart has been changed from the earlier examples, in
order to also show the ground terrain.
Example
A complete picture of road construction showing cuts and fills, roadworks,
fences etc. may be put on a time-chainage chart, for example Figure 7.22.
(b)
(a)
5 days
5 days
5 days
5 days
200
500
2 6 7 8 11
Days
(c)
S/S lead time
period 4 days
F/S lead time
period 1 day
Cut work
Drainage work
Drainage work
Drainage work
Cut work
Cut work
C
h
a
i
n
a
g
e

(
m
)
Figure 7.20 Earthmoving example. (a) Bar chart with S/S relationship; (b) Bar chart
with F/S relationship; (c) Time-chainage chart.
0
200 400 600
Chainage (m)
Days
0
3
6
9
Terrain
Drains 1 work
Drains 2 work
Drains 3 work
Figure 7.21 Drain construction example.
254 Conventional treatment from a systems perspective
Cut
Fill
Subbase
Seal,
surface
Furniture
Landscaping, fencing
Culvert
Subbase
Seal,
surface
Furniture
Clear
Jan
Feb
Mar
Apr
May
Chainage
Fill
Original ground line
Culvert
Finished road line
Cut
Figure 7.22 Time-chainage chart for road construction. (Activity names have been
abbreviated because of space limitations – work items are implied.)
Similar diagrams can be used for railway construction, pipeline construc-
tion and so on.
Case example – Aircraft engines
In an organisation involved in the repair and overhaul of aircraft
engine turbine blades and vanes, the operations or production work
is divided into two main sections, namely:

Manufacturing Operations

Special Process Operations.
The Manufacturing Operations section is divided into various product
cells, namely the high pressure turbine blades, high pressure turbine
vanes, low pressure turbine blades and low pressure turbine vanes.
Each product cell is arranged in a U-shaped formation. The principle
behind this production concept is to have as many operations as
possible located within the cells to cut down on operator and parts
movement, which are considered unnecessary and hence labelled as
Selected topics 255
waste. The idea is to have the parts move from one station to another
in a ‘pull system’. That means output is only what is required by the
next customer or next operation.
In earlier days, the production systemwas similar to a ‘push system’,
whereby product was pushed through the process in batches, and
quantities were large enough to meet present and future customer
demand and compensate for identified and unidentified problems in
the manufacturing process. The push system had no discipline on
inventory, but regarded more as always better.
In the ‘push system’, processes were broken up into specific groups,
for example welding, blending, machining and grit-blasting. With the
implementation of the ‘pull system’, all common equipment was sep-
arated and placed into each one’s respective product cell. At the end
of the day, more equipment was required in order to have a complete
self-sustaining cell with smooth movement of product from station-
to-station. What this meant was that the purchasingof additional equip-
ment didnot have togothroughanyproper justificationstudy. The only
reason given was that the cell needed it. Utilisation was never an issue.
The Special Process Operations consisted mainly of batch processes
undergoing for example, brazing, coating, plasma spraying and heat
treatment. The ‘pull system’ was avoided. The justification for pur-
chasing is quite different to that of the Manufacturing Operations
section. Because most of the equipment in the Special Process Opera-
tions section is highly capital intensive, a thorough feasibility study is
required. The selection of equipment is based on technical as well as
economic considerations. Utilisation is the most important criterion.
The company cannot afford to have an expensive ‘white elephant’
sitting around, if it can be avoided.
Continuous components
Time-chainage charts
Time-chainage charts are seen to provide some information additional to
that on connected bar charts. Time-chainage charts contain lines and boxes.
Within boxes no other activities may intrude. The charts show:

The sequence of activities over time

The location of activities in space

The direction of production/work

Activity production rate

Buffers (intervals) between activities

Float.
256 Conventional treatment from a systems perspective
Example
Consider a pipelaying operation which involves the following major activ-
ities (Carmichael and Bird, 1992):
Clear and Grade
Excavation
Stringing
Welding
Backfilling
Restoration.
The activity Clear and Grade involves the clearing of the pipe route of
obstacles for construction (clears trees and scrub, opens fences, installs
temporary gates, strips and windrows top soil for later spreading during
restoration) and establishing acceptable construction working grades.
Excavation of the trench follows the Clear and Grade activity. Typically
the trench will be, say, 300 mm wider than the diameter of the pipe and
provide for cover over the pipe. The bottom of the trench is bedded in sand
in preparation for receiving the pipe.
Stringing the pipe follows trenching; the possibility of damage to the
insulating coating applied to the pipe is an important consideration.
The important Welding operation follows the stringing. The Welding
operation includes the radiographic testing of the welds to prove their
integrity prior to joint coating and lowering the pipe into the ground.
When the welds have been cleared and the joints coated, the pipe is
lowered into the trench and the trench backfilled. Backfilling sees fine-
grained padding material placed around the pipe, and then the return of
the excavated material.
Restoration of the pipe route proceeds after backfilling and results in the
disturbed areas being returned to as near their original undisturbed contours
as possible. Top soil windrowed during the clearing and grading is spread
across the right of way and the brush and timber cut during clearing is also
spread across the work area to help stabilise and promote rapid regeneration.
Pipeline construction can be likened to a production line, the only dif-
ference being that in this case the ‘factory’ moves. This movement may be
several kilometres per day.
Each of the above activities is carried out sequentially and in order that a
disruption does not stop the whole production train, separations (buffers)
equivalent to a few day’s production are maintained between activities.
As each section of the pipeline is backfilled and restored, the pipe is
pressured to test both the structural strength and leak tightness. The pipe
is then commissioned and handed over to the operating organisation.
Using a network diagram, the activities would be represented as in
Figure 7.23a. Should there be concern about the durations to perform these
Selected topics 257
Stringing Welding Backfilling Restoration
Clear
and
grade
Excavation
(a)
(b)
Stringing
Welding
Clear
and
grade
Excavation
Backfilling Restoration Buffer
Buffer Buffer Buffer
Buffer
Figure 7.23 Network representation of pipelaying project. (Activity names have been
abbreviated because of space limitations – work items are implied.)
activities, then buffers may be inserted into the diagram as in Figure 7.23b.
The buffers are contingencies and allow for disruptions. For this example, it
is estimated that each of these activities will take 20 weeks and start 1 week
after the previous activity. An allowance of 2 weeks is made to separate fol-
lowingactivities. Progress is targetedfor 3kilometres per week, withacycle of
4 weeks work followed by 1 week leave. This information is transferred onto
Figure 7.24 which has axes of cumulative production and time.
Figure 7.24 assumes that production for each of the activities progresses
at the same constant rate. Should this not be the case, then the lines for each
activity will not be parallel (for example, Figure 7.25). The cumulative pro-
duction plots then assist in working out the required resource numbers for
each activity in order that following activities are not interfered with, while
at the same time ensuring the necessary target production. This is a com-
mon way that cumulative production plots are developed – the estimated
production rate for each activity is used to define the slope of that activity’s
line on the chart. It is the reverse of the procedure as illustrated in Figure
7.25 where the slope of the activity’s line establishes what production and
hence resource numbers are needed for that activity.
Discrete components
Cumulative production plots take on a slightly different appearance where
discrete components are involved, rather than something continuous like
a pipeline, railway or road. For a bar chart such as Figure 7.26a, joining
258 Conventional treatment from a systems perspective
Clear
and
grade
5 10 15 20 25 30 35 40
Weeks
K
i
l
o
m
e
t
r
e
s
50
40
30
20
10
Leave
Restoration
Figure 7.24 Schedule determined by production (parallel schedule). Activities of
Excavation, Stringing, Welding and Backfilling not shown for clarity.
(Activity names have been abbreviated because of space limitations –
work items are implied.)
the bars (that is, ‘integrating’ the bar chart) and changing the vertical axis
gives the cumulative production plot of Figure 7.26b. All the work for each
discrete component (in this case a floor of a building) is located on the one
line across the page.
When‘integrated’, abar chart becomes the same as atime-basedcumulative
productionplot. Cumulativeproductionplots, bar charts andmultipleactivity
charts contain essentially the same information. It is the additional vertical
scale that makes the cumulative production plot attractive, compared to a bar
chart. The computational path to the cumulative production plot should be
no different to that to a conventional bar chart, namely via a network.
Example
An alternative drafting form is illustrated in the following example. The
simplified example involves the construction of a bridge represented by
three activities – place piers, place beams and place slab – with the sequence
repeated tenfold to finish the bridge. Durations for each of these three
activities are 9 days, 6 days and 5 days respectively based on using one
team of workers for each activity. With buffers of 2.5 days between the
Selected topics 259
Clear
and
grade
5 10 15 20 25 30 35 40
Weeks
K
i
l
o
m
e
t
r
e
s
50
40
30
20
10
Restoration
Figure 7.25 Schedule determined by resource numbers (resource schedule).
(Activity names have been abbreviated because of space limitations –
work items are implied.)
first and second activities and between the second and third activities, this
gives a total duration of 25 days for each of the 10 bridge spans.
Assuming that the target completion duration of the total bridge is 47.5
days and each of the three activities is to be carried out at the same pro-
duction rate, then the cumulative production plot of Figure 7.27a follows.
Resource numbers are acquired to achieve the planned activity production
rates. From Figure 7.27b four teams of workers are required for the pier
work, three teams of workers are required for the beam work and two
teams of workers are required for the slab work.
The alternative approach to the scheduling is via the resource schedule
(schedule determined by resource numbers) rather than the parallel schedule
(schedule determined by production). Here the resource numbers available
are first determined and the corresponding activity production rates that
can be achieved with these available resource numbers are then calculated.
For example, assume that in the placing of the piers, two teams are
employed. The first team would start on the first bridge span on day 11.5
and finish on day 17.5. The second team could then undertake work on
the second bridge span. The third span is due to start on day 16.5 but the
first team is not available until day 17.5, completing the third span on day
(a)
G
F
l
o
o
r
/
s
t
o
r
e
y
M
1
2
3
4
5
6
7
8
9
P/R
R
Days
I
n
t
e
r
n
a
l

p
a
r
t
i
t
i
o
n
s
R
o
u
g
h
-
i
n

s
e
r
v
i
c
e
s
a
n
d

w
e
t

a
r
e
a
s
R
o
u
g
h
-
i
n

s
e
r
v
i
c
e
s
a
n
d

w
e
t

a
r
e
a
s
F
i
n
i
s
h
e
s
12
10
11
0
1
2
3
4
5
6
7
8
9
Days
F
l
o
o
r
s
I
n
t
e
r
n
a
l

p
a
r
t
i
t
i
o
n
s
F
i
n
i
s
h
e
s
(b)
Figure 7.26 Example construction program for internal work for a building. (a) Bar chart;
(b) Cumulative version. (Activity names have been abbreviated because of
space limitations – work items are implied.)
40 60
0
2
4
6
8
10
20 0
Time (days)
S
p
a
n
Pier
(a)
Beam
Slab
Target completion
time = 47.5 days
20 40 60
0
Time (days)
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
Pier
Beam
Slab
(b)
Pier team 2
Pier team 3
Pier team 4
Beam team 2
Beam team 3
Pier team 1
Slab team 1
Slab team 2
Beam team 1
Figure 7.27 Bridge example. (a) Schedule determined by production (parallel sched-
ule). (b) Bar chart for parallel schedule. (Activity names have been
abbreviated because of space limitations – work items are implied.)
262 Conventional treatment from a systems perspective
23.5. The same constraint applies to the second team. That is, the cycle
is 23.5−17.5 = 6 days per two spans. Or on average, one span (piers) is
completed every 3 days. This is shown in Figure 7.28a. (By comparison
Figure 7.27 implies a rate of finishing one span (piers) every 2.5 days.)
Figure 7.28b shows the same information in bar chart form.
0
2
4
6
8
S
p
a
n
10
20 40 60
0
Time (days)
Pier
(a)
Beam
Slab
Target completion
time = 52 days
20
40 60
0
Time (days)
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
Pier
Beam
Slab
(b)
Pier team 1
Pier team 2
Pier team 3
Pier team 4
Beam team 1
Beam team 2
Slab team 1
Slab team 2
Figure 7.28 Bridge example. (a) Schedule determined by resource numbers (resource
schedule). (b) Bar chart for resource schedule. (Activity names have been
abbreviated because of space limitations – work items are implied.)
Selected topics 263
Should three teams be employed to place the piers, the following sit-
uation develops. Going backwards in time, the last span of the bridge
can be finished on day 40 and started on day 34. That is, the fourth
last (or seventh) span can finish on day 34 and start on day 28. (And so
on for the seventh last span.) That is, the cycle is 40 −34 = 34−28 = 6
days per three spans. Or on average, one span (piers) is completed every
2 days. This is shown in Figure 7.29a. Figure 7.29b shows the same
information in bar chart form. Interestingly, increasing the team num-
bers from two to three has no effect on the overall project completion
time because the placing of the piers still has to fit in with the other two
activities.
Replanning
Replanning can be assisted by overplotting on the cumulative production
plots, actual work progress as indicated in Figures 7.30 and 7.31. The rate
of production and potential conflict between activities is perhaps not so well
seen on bar charts. Controls may be chosen so as to prevent interference
between activities.
0
2
4
6
8
S
p
a
n
10
20 40 60
0
Time (days)
Pier
(a)
Beam
Slab
Target completion
time = 52 days
Figure 7.29 Bridge example. (a) Resource schedule; (b) Bar chart for resource sched-
ule. (Activity names have been abbreviated because of space limitations –
work items are implied.)
20
40 60
0
Time (days)
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
1
2
3
4
5
6
7
8
9
10
Pier
Beam
Slab
Slab team 1
Slab team 2
(b)
Beam team 3
Beam team 2
Beam team 1
Pier team 1
Pier team 2
Pier team 3
Pier team 4
Figure 7.29 (Continued).
Clear
and
grade
5
10 15 20 25 30 35 40
Weeks
50
40
K
i
l
o
m
e
t
r
e
s
30
20
10
Restoration
Actual
As-planned
Figure 7.30 Scheduled and actual work; example. (Activity names have been
abbreviated because of space limitations – work items are implied.)
Selected topics 265
Excavate trench
Lay pipe
20 days
Lay pipe
0
20
Days
12 days
8
Excavate
trench
C
h
a
i
n
a
g
e
Time
now
As-planned
Actual
F/F lead time
period 0 days
Figure 7.31 As-planned and actual work; example.
Reference
Carmichael, D. G. and Bird, T. R. (1992), Project Management and
Engineering – An Introduction, Engineering Education Australia (EEA),
Sydney.
Exercises
1. Using a similar format to Figure 7.26, what would the program look
like where a building was being demolished floor by floor, starting at the
top of the building?
2. In repetitive construction, it is important to provide continuity of work
for people and plant. For example in multistorey building construction it
is important that the different trades when finished working on one storey
have work to start on the following storey.
How does the cumulative production plot ensure that continuity of work
is obtained?
3. How do you alter a cumulative production plot in order to reduce
non-productive periods?
4. For an activity represented in a cumulative production plot by a sin-
gle line at 60

to the vertical, how does the shape of the line change if,
whenever the work is repeated, a 5% improvement occurs in the duration
taken to undertake that work? Sketch graphically what is happening to the
cumulative production plot.
Part II
Synthesis treatment
Chapter 8
Multistage planning
Introduction
Planning over the total project duration may be simplified by considering
the project to be composed of stages (or phases). The planning problem
then becomes a multistage optimal control problem or multistage decision
problem. Various optimal control tools are available to assist, including
Pontryagin’s maximum principle, dynamic programming and mathematical
programming.
Controls are chosen for each stage to give a globally optimum solution
over the whole project. The controls considered in this chapter relate to
resources and resource production rates only, that is the lower level prob-
lems in the context of Chapter 5. The work method is assumed given or
fixed. The synthesis problem involving work method as controls is very
difficult; in the comparative area of structural optimisation, only Michell
structure results have evolved.
States are chosen as cumulative resource (or money) usage and cumulative
production.
The single objective case is given.
Stages may be chosen to match the type of work, convenient time period
divisions in the project, or in an arbitrary sense. Hold or approval points
might be used to distinguish stages. Breaking down the project into stages
is equivalent to discretising the project duration interval.
This chapter is dealing with planning over the whole project duration
using stages. Replanning is also carried out according to stages or at discrete
points in time, but this follows from the nature of what replanning is.
Parts of this chapter follow Carmichael (2004).
Stage modelling
Introduction
A (planned) baseline is obtained by solving the deterministic optimal control
problemfromproject start to project end (or the equivalent iterative analysis
270 Synthesis treatment
version). Future uncertain events are converted into equivalent deterministic
influences. (There is no need to talk of risk management.) This optimal
control problem may be expressed in continuous time or discrete time. The
latter is treated here.
Planning over the total project duration
A project may be decomposed into stages. For sequential stages (that is,
the staging is based on a time period subdivision; each stage represents a
period of time) the decomposition will look like Figure 8.1. Each stage is a
subproject.
In Figure 8.1,
N number of stages
l stage counter, l =0, 1, 2, . . . , N−1
x(li state leading into stage l
x(0i initial state
x(Ni final state
u(li control at stage l
State and control are vector functions, x = (x
1
, x
2
, . . . , x
n
i
T
, u =
(u
1
, u
2
, . . . , u
r
i
T
Associated with each stage is an objective function J(li, l =
0, 1, 2, . . . , N –1. (See the earlier definition of the term ‘objective’ used in
this book. It is not used in a loose lay person, dictionary or management
text sense.)
Stage 0 Stage 1
Stage 2 Stage N–1
u(0)
u(0)
u(1)
u(1)
u(2) u(N–1)
u(N–1)
. . .
x(0)
x(0)
x(1)
x(1)
x(2)
x(2)
x(3) x(N–1)
x(N–1)
x(N)
x(N)
Stage 0 Stage 1 Stage N–1
. . .
(a)
(b)
Figure 8.1 Staged (phased) decomposition of a project. (a) ‘Activity-on-link’ form;
(b) ‘Activity-on-node’ form.
Multistage planning 271
Note that the stage output and the state are the same in this stripped-
back project formulation.
Such a formulation is amenable to the application of discrete opti-
mal control systems theory techniques such as dynamic programming.
There, an optimisation problem is broken down into subproblems asso-
ciated with each stage. The optimisation problem is solved by succes-
sively going through each stage, from start to finish (or finish to start).
Associated with each stage is an objective function which evaluates the
worth of each stage control selected from a range of possible alternative
controls. The state represents the connection between successive stages;
it carries forward information on the system behaviour, such that the
state x(l) incorporates information from stages 0 to l–1.
What constitutes the state and control is perhaps the hardest thing
to establish when setting up such an optimal control formulation. The
state differs from application to application.
The state at any stage is the minimal amount of information needed
to completely determine the behaviour (state) of the system for all
other following stages, for any given control. To extend the concept
of state to stochastic systems, the state at any stage is regarded as
the information that uniquely determines the probability distributions
of behaviour (state) at all other following stages. That is, a discrete
equivalent of a Markov process.
(Carmichael, 1981)
The state is information needed from all preceding stages in order
to select a control in the current stage without reference to previous
controls selected. It represents status information on the project at any
point in time.
(Carmichael, 2004)
In line with Chapter 2, the following are selected as the states and controls.
(The work method is assumed fixed.)
state(l): cumulative resource usage; the resource usage of stages up
to and including stage l–1
cumulative production; the production of stages up to and
including stage l–1
control(l): resource numbers chosen at stage l (for each resource type)
production chosen at stage l
In terms of the resource component of the above state and control defi-
nitions, cumulative resource usage is in fact the representation of S curves
272 Synthesis treatment
Resource usage
M
o
n
e
y
Figure 8.2 Some possible relationships between resource usage and money.
(resources), while the number of states and controls will reflect the number
of resource types.
This makes for a very high dimensional problem. The dimension can
be reduced if resources are expressed in a common unit of money and
aggregated (Figure 8.2).
state(l): cumulative cost; the expenditure of stages up to and includ-
ing stage l–1
cumulative production; the production of stages up to and
including stage l–1
control(l): expenditure chosen at stage l
production chosen at stage l
In terms of cost, cumulative expenditure is that plotted in the familiar S
curve (cost).
x(li and u(li are vectors with the above listed components.
State equations
Given the above definitions of state and control, the system equations
(model) in state equation (state transformation) form become
x(l+1i =x(li +u(li
where
x(li state for stage l
u(li control in stage l
Multistage planning 273
k
x(k)
Figure 8.3 Example state trajectories.
Time
P
r
o
d
u
c
t
i
o
n
Lines of different
resource performance
Figure 8.4 Example ‘constitutive relationships’ for a resource. The slope of a line is
a (resource) production rate.
The initial state x(0i =0, but the final state x(Ni may not be fixed.
The above state equations are characteristic of discrete (data) control
systems.
State trajectories may look something like Figure 8.3.
To track project schedule, the constituent level behaviour of Figure 8.4
is used.
Linearity
Because of the linear state equations, the optimal solution may turn out to
be singular, possibly even bang-bang in nature (Carmichael, 1981). This
will depend on the nature of the objective and constraints chosen.
274 Synthesis treatment
Objectives and constraints
Some example objectives and constraints are given below.
Objectives
There are a number of possible choices of objectives available to the plan-
ner. The most obvious would be minimum resource usage/expenditure and
minimum duration.
Minimum resource usage/expenditure. For each stage, J(li =u(li and
minJ =min
N−1
¸
l=0
u(li
where only the components of u relating to resources are involved.
Minimum duration. For each stage, J(li =T(li, the stage duration. And
minJ =min
N−1
¸
l=0
T(li
Assuming the constituent level behaviour of Figure 8.4 holds, minimum
duration will correspond to maximum production.
minJ =max
N−1
¸
l=0
u(li =min−
N−1
¸
l=0
u(li
where only the component of u relating to production is involved.
Constraints
Typically, there may be constraints on the overall expenditure and project
duration,
x(Ni ≤some limit
where only the component of x relating to cost is involved, and
N−1
¸
l=0
T(li ≤some limit
or
N−1
¸
l=0
u(li ≥some limit
Multistage planning 275
where only the component of u relating to production is involved.
There may also be restrictions on resource usage/expenditure throughout
the project,
u(li ≤some limit l =0, 1, 2, . . . , N–1
where only the components of u relating to resources or expenditure are
involved.
Optimisation problem
A generalisation
The above development may be shown to fit within a general optimal con-
trol problem statement (Carmichael, 1981). This section is after Carmichael
(2004).
System model
The general form of the system (state) equations given in the previous
section is
x(l+1i =F|x(li, u(li, l] l =0, 1, . . . , N–1
where
x(li = |x
1
, . . . , x
n
]
T
is an n-valued vector of the state at stage l
u(li = |u
1
, . . . , u
r
]
T
is an r-valued vector of controls at stage l
F = |F
1
, . . . , F
n
]
T
is a nonlinear vector function.
For given u, and initial and final (terminal) conditions on x, x is completely
defined by the state equations.
The equations represent a sequence of transitions from the lth to the
(l+1ith state, l = 0, 1, . . . , N–1. The state is assumed to have Markov
properties; that is, the state at l+1 depends only on the immediately previous
state x(li and control u(li.
Remaining optimal control problem components
Objective
A general objective, incorporating the earlier examples, can be written as,
minJ = g|x(li]
l=N
l=0
+
N−1
¸
l=0
G|x(li, u(li, l]
276 Synthesis treatment
where
G, g are scalar single-valued functions of their respective arguments.
Constraints
Constraints influence the solution by isolating admissible solutions from all
possible solutions. Constraints may be given in the form of equalities or
inequalities. Constraints may exist on both state and control variables.
Common project constraints include limits on expenditure and limits on
resource usage.
The problem
The problem is to determine an admissible sequence of controls ˆ u(li, l =
0, 1, . . . , N–1 satisfying the system (state) equations, initial and final (ter-
minal) conditions on the state, and any constraints, and minimising J. ˆ u(li,
l =0, 1, . . . , N–1 is termed the optimal sequence or policy.
The linear–quadratic case
The case involving linear system equations and quadratic objective has a
particularly neat solution (Carmichael, 1981).
Singular control
Where the system (state) equations and objective are linear, the solution
may be singular or bang-bang in nature.
Determinism
The above is a deterministic formulation. A related stochastic formulation
can be given.
Examples
Introduction
Two examples are given below. The first chooses project staging such that
the Markov assumptions in the state hold, and this leads to the canonical
system equations.
x(l+1i =F|x(li, u(li, l] l =0, 1, . . . , N–1
Multistage planning 277
Relevant solution techniques are Pontryagin’s maximum principle,
Bellman’s dynamic programming and mathematical programming
(Carmichael, 1981).
With a linear objective and constraints on the controls, bang-bang and
singular solutions may emerge.
The second example has arbitrarily chosen staging. System equations can
still be derived but do not follow the canonical form. Here mathematical
programming would be the preferred optimisation technique.
Canonical form example
Consider the example network shown in Figure 8.5(a,b). The work method
leading to the network is assumed fixed. That is, the project level control
is fixed or given.
The project may be subdivided into subprojects or stages. The subdi-
vision which preserves the Markov assumption on the state is given in
Figure 8.6(a,b). The lines through the networks in Figure 8.6(a,b) indicate
the divisions between subprojects or stages.
Controls u(t, j i are shown adjacent to each activity (t, j i in the activity-
on-link diagram, and in the activity boxes in the activity-on-node diagram.
Using the notation xj as the state following activity (t, j i, the state in and
state out (cumulative resource usage/expenditure, cumulative production)
for each stage are as follows:
x1 given
x4 = x1+u(1,2i +u(2,4i +u(1,3i +u(3,4i
x9 = x4+u(4,5i +u(5,7i +u(4,6i +u(6,7i +u(7,8i +u(8,9i
+u(6,9i
x12 = x9+u(9,11i +u(9,10i +u(10,11i +u(11,12i
These equations belong to the standard (canonical) form
x(l+1i =F|x(li, u(li, l] l =0, 1, . . . , N–1
where
N =3
x(0i =x1
x(1i =x4
x(2i =x9
x(3i =x(Ni =x12
u(0i =u(1,2i +u(2,4i +u(1,3i +u(3,4i
u(1i =u(4,5i +u(5,7i +u(4,6i +u(6,7i +u(7,8i +u(8,9i +u(6,9i
u(2i =u(9,11i +u(9,10i +u(10,11i +u(11,12i
Concept
design
Feasibility
Documents
Tender
review
Tender
Familiarisation
Data
compilation
Submission
Negotiation
Tender
modification
Tender
evaluation
Team
assembly
Team
mobilisation
Approvals
Detailed
design
Concept
design
Detailed
design
Feasibility Documents
Tender
review
Tender
Familiar-
isation
Data
comp’ln
Subm’n
Negot’n
Tender
modif ’n
Tender
evaluat’n
Team
assembly
Team
mobil’n
Approvals
(a)
(b)
Figure 8.5 Example networks. (a) Activity-on-link diagram; (b) Activity-on-node diagram. (Activity names have been abbreviated
because of space limitations – work items are implied.)
Multistage planning 279
1
2
3
4 5
6
7 8
9
10
11 12
u(1,2)
u(1,3)
u(2,4)
u(3,4)
u(4,5)
u(5,7)
u(4,6)
u(6,7)
u(7,8)
u(6,9)
u(8,9)
u(9,10)
u(9,11)
u(10,11)
u(11,12)
Stage 0 Stage 1 Stage 2
(a)
(b)
u(1,2)
u(1,3)
u(2,4)
u(3,4)
u(4,5)
u(4,6)
u(5,7)
u(6,7)
u(6,9)
u(7,8)
u(8,9)
u(9,10)
u(9,11) u(11,12)
u(10,11)
x2
x3
x4
x5
x6
x7 x8
x9
x10
x11
x12
x1
Stage 2 Stage 1 Stage 0
Figure 8.6 Example staging of project. (a) Activity-on-link diagram; (b) Activity-on-node
diagram.
Non-canonical form example
Consider the example network shown in Figure 8.7(a,b). The work method
leading to the network is assumed fixed. That is, the project level control
is fixed or given.
The project may be subdivided into subprojects or stages. One subdivision
has been arbitrarily chosen and is given in Figure 8.8(a,b), but the following
development applies to any other similar subdivision. The lines through the
networks in Figure 8.8(a,b) indicate the divisions between subprojects or
stages; these lines do not have to be vertical. In the activity-on-node dia-
gram of Figure 8.8b, the dummy activities of the activity-on-link diagram,
Figure 8.8a, have been added to show the comparable development.
Controls u(t, j i are shown adjacent to each activity (t, j i in the activity-
on-link diagram, and in the activity boxes in the activity-on-node diagram.
Using the notation xj as the state following activity (t, j i, the state in and
state out (cumulative resource usage/expenditure, cumulative production)
for each stage are as follows:
280 Synthesis treatment
Assemble
options
Apply
finishings
Marketing
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Dummy
Dummy
(a)
(b)
Assemble
options
Apply
finishings
Marketing
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Figure 8.7 Example networks. (a) Activity-on-link diagram; (b) Activity-on-node diagram.
(Activity names have been abbreviated because of space limitations – work
items are implied.)
Stage 0
x1 x3
x3 =x1+u(1, 2i +u(2, 3i
Stage 1
x6
x3 x7
x5
x6 =x3+u(3,4i +u(4,6i
x7 =x3+u(3,4i +u(4,7i +x3+u(3,5i +u(5,7i
Multistage planning 281
(a)
(b)
Stage 0 Stage 1 Stage 2
1
2 3 4
5
6
7
8 9
10
11 12
u(1,2)
u(2,3)
u(3,4)
u(3,5)
Stage 3
u(4,6)
u(4,7)
u(5,7)
u(5,11)
u(7,9)
u(8,9)
u(9,10)
u(10,11)
u(11,12)
u(6,8)
Stage 0 Stage 1
Stage 2
Stage 3
u(1,2)
u(2,3) u(3,4) u(4,6) u(6,8) u(8,9)
u(3,5) u(7,9)
u(5,11)
u(9,10) u(10,11)
u(11,12)
x1
x2
x3 x4 x6 x8
x9
x10
x11
x12
x3
x4
x5
x5
x9
x11
u(4,7)
u(5,7)
x7
x7
Figure 8.8 Example staging of project. (a) Activity-on-link diagram; (b) Activity-on-node
diagram.
Accounting for redundancies
x7 =x3+u(3,4i +u(4,7i +x3−x3+u(3,5i +u(5,7i
x5 =x3+u(3,5i
Stage 2
x6 x10
x7
x5 x5
x5 =x5
x10 =x6+u(6, 8i +u(8, 9i +u(9, 10i +x7+u(7, 9i +u(9, 10i
282 Synthesis treatment
Accounting for redundancies
x10 =x6+u(6,8i +u(8,9i +u(9,10i +x7−x3−u(3,4i +u(7,9i
+u(9,10i −u(9,10i
Stage 3
x10
x5
x12
x12 =x10+u(10,11i +u(11,12i +x5+u(5,11i +u(11,12i
Accounting for redundancies,
x12 =x10+u(10,11i +u(11,12i +x5−x5+u(5,11i +u(11,12i
−u(11,12i
These system equations may be collectively written as,








x3
x5
x6
x7
x10
x12








=








x1
x3
x3
x3
x6+x7−x3
x10








+








u(1,2i +u(2,3i
u(3,5i
u(3,4i +u(4,6i
u(3,4i +u(3,5i +u(4,7i +u(5,7i
u(6,8i +u(8,9i +u(9,10i +u(7,9i −u(3,4i
u(10,11i +u(5,11i +u(11,12i








with initial conditions x1 given.
The system equations are linear. Controls can be grouped, for example, as
they appear in the above system equations, as a tidying up exercise to give,








x3
x5
x6
x7
x10
x12








=








x1
x3
x3
x3
x6+x7−x3
x10








+








u1
u2
u3
u4
u5
u6








Multistage planning 283
where,
u1 =u(1,2i +u(2,3i
u2 =u(3,5i
u3 =u(3,4i +u(4,6i
u4 =u(3,4i +u(3,5i +u(4,7i +u(5,7i
u5 =u(6,8i +u(8,9i +u(9,10i +u(7,9i −u(3,4i
u6 =u(10,11i +u(5,11i +u(11,12i
But other options are available for grouping.
The dimension of these system equations can be reduced by reducing the
number of stages, and carefully selecting the places where stage divisions
occur.
The solution to the associated optimisation (planning) problem might
best be performed using mathematical programming.
Chapter 9
Synthesis over levels
Direct synthesis approach
Much of the iterative nature can be removed from the planning process,
if a synthesis approach is adopted. The essential difference between
the analysis-based and synthesis-based procedures is in the level of
abstraction adopted in the computations. (The terminology ‘level of
abstraction’ is used in the sense relating to the quantity of a priori data
assumed.) Analysis-based techniques impose a total project configura-
tion ab initio, while the plan emerges fromany given level of abstraction
as a natural consequence of the direct synthetic treatment. Presumably
the extreme generality that may be attained in the direct case would
involve little or no a priori knowledge of the emerging project. How-
ever, for a solution of practical significance, certain leading properties
of the project are best assumed. The choice of abstraction level on
which the planner chooses to work would be a balance between his/her
expertise-based judgements and desired computation load. A synthesis-
type treatment can only proceed where certain of the project properties
remain free and adjustable. A synthesis-type format to planning is the
fundamental approach.
(Carmichael, 2004)
The synthesis approach selects the most appropriate controls from the
range of possible controls. The selection is determined by the objectives and
constraints.
Within a hierarchy of levels, a problem on one level is nested within the
next higher level problem.
Models
When modelling projects, as with most systems, there exists an interdepen-
dence among levels. No isolated levels exist. A model at any level contains
lower level information, that is, lower level information transfers to higher
levels.
Synthesis over levels 285
The following modelling applies to stripped-back projects, as defined
earlier.
Constituent model
The constituent level is the basic behaviour level of a resource (person or
piece of equipment).
The model at the constituent level is the resource equivalent to a material
technologist’s constitutive relationship. Each resource will have a different
‘constitutive relationship’, as will a resource performing under different
conditions. For definiteness, the term ‘production’ is used here, meaning
some measure of work done. A linear example in terms of production is
shown in Figure 9.1; nonlinear and integer relationships are possible. The
measurement unit of production can be various, for example m
3
, m or item.
Schematically, Figure 9.2 applies.
For each resource type,
Resource production =Resource production rate ×Time period
and
Resource unit production =Resource production rate ×1 = H
where H is the resource production rate.
Time
Production
Lines of different
resource performance
Figure 9.1 Example ‘constitutive relationships’ for a resource. The slope of a line is
a (resource) production rate.
Constituent
resource
(Resource)
production
(Resource)
production rate
(and time period)
Figure 9.2 Inflow–outflow relationship at the constituent level.
286 Synthesis treatment
The control at the constituent level is H for each resource type; resource
behaviour is given by its production.
Element model
The inflow–outflow relationship of an element of an activity, in production
terms, might be represented as in Figure 9.3. An element represents 1 unit
of time, for example 1 day or 1 hour.
For each resource type,
Element production =Element resource number ×Resource unit
production
= µH
where, µ is the resource number (quantity), µ =1, 2, . . .
If the different resource types have the same unit of measurement for
production, the element productions for all resource types can be added to
give a total element production.
Figure 9.4 shows the influence of resource numbers (quantities) on pro-
duction.
In resource usage terms, an element uses resource numbers of µ (for each
resource type).
The controls at the element level are µH and µ for each resource type; the
behaviour is given by element production and resource number (quantity)
for each resource type.
Resource numbers and type can be measured or expressed in money
terms.
Activity model
An activity of duration I (for example I days or I hours) is composed of
I elements each of duration 1 (for example 1 day or 1 hour).
In terms of production, individual elements as in Figure 9.5 combine to
give an activity as in Figure 9.6. Typically, the inflow from, and outflow
Element
Resource
unit production
Element
resource number
Element
production
Figure 9.3 Element of an activity; for each resource type.
Synthesis over levels 287
Time
Production
3
4
5
2
1
Lines of equal
resource numbers
Figure 9.4 Example resource production rate plot for different resource numbers.
Element e
Element production
Cumulative element
production
(from preceding element)
Cumulative element
production
(to following element)
Figure 9.5 Individual element with element interaction.
. . .
Element
e = 1
Element
e = 2
Element
e = E
Activity a
Element
production
Element
production
Element
production
Activity
production
Figure 9.6 Activity; for each resource type.
to, other elements will be cumulative element production, and the control
will be element production.
In Figure 9.6, the arrows between elements represent cumulative
production.
In Figure 9.5, the element interaction gives, for each resource type,
Cumulative element production outflow
=Element production+Cumulative element production inflow
= µ+Cumulative element production inflow
288 Synthesis treatment
In Figure 9.6, the activity inflow–outflow relationship is, for each resource
type,
Activity production =Sum of element productions
=
¸
All elements
Element productions =
¸
All elements
µH
In terms of resource usage, individual elements as in Figure 9.7 combine to
give an activity as in Figure 9.8. Typically, the inflow from, and outflow
to other elements will be cumulative element resource numbers, and the
control will be element resources (number and type). Resource numbers can
vary from element to element, to suit the activity; there is not a requirement
for uniform resource numbers.
In Figure 9.8, the arrows between elements represent cumulative element
resource numbers.
In Figure 9.7, the element interaction gives, for each resource type,
Cumulative element resource number outflow=Element
resource number +Cumulative element resource number inflow
= µ+Cumulative element resource number inflow
Element e
Cumulative element
resource number
(from preceding element)
Cumulative element
resource number
(to following element)
Element
resource number
Figure 9.7 Individual element with element interaction.
. . .
Element
e = 1
Element
e = 2
Element
e = E
Activity a
Activity
resource
number
Element
resource
number
Element
resource
number
Element
resource
number
Figure 9.8 Activity composed of elements; for each resource type.
Synthesis over levels 289
In Figure 9.8, the activity inflow–outflow relationship is, for each resource
type,
Activity resource number = Sum of element resource numbers
=
¸
All elements
Element resource numbers
=
¸
All elements
µ
The controls at the activity level are
¸
All elements
µH and
¸
All elements
µ for each resource
type; the behaviour is given by activity production and activity resource
number (quantity), for each resource type.
Resource numbers and type can be measured or expressed in money
terms. For the special issue of reducing the duration to perform (compress)
an activity, the total resource usage may increase, and the cost may increase
as in Figure 9.9.
Project model
Overlapping relationships between activities may be taken care of through
using subactivities which are themselves activities. In aggregated form,
Figures 9.6 and 9.8, together with interaction between activities, convert to
Figures 9.10(a,b) respectively.
For each resource type, for each activity, these can be expressed as, for a
production focus,
Cumulative activity production outflow=
¸
All elements
Element productions
+Cumulative activity
production inflow
Activity duration
D
i
r
e
c
t

c
o
s
t
Figure 9.9 Changing activity cost with activity duration.
290 Synthesis treatment
Activity a
Cumulative
activity production
(from preceding activities)
Cumulative activity
resource numbers
(from preceding activities)
Cumulative
activity production
(to following activities)
Σ Element productions
Activity a
Cumulative activity
resource numbers
(to following activities)
Σ Element resource numbers
(a)
(b)
Figure 9.10 Aggregated activity representation. (a) Production focus; (b) Resource
usage focus.
For each resource type, for each activity, for a resource usage focus,
Cumulative activity
resource numbers outflow
=
¸
All elements
Element resource numbers
+Cumulative activity resource numbers
inflow
Activities can be assembled to give a project (or . . . , work package, sub-
subproject or subproject). For example, consider Figure 9.11.
Activity inflow and outflow will reflect the project network connections,
but redundancy of resource and production information through parallel
network paths has to be acknowledged. That is, for each activity,
Activity inflow=Other (preceding) activities outflow
(ignoring any redundant accounting)
Activity outflow=Activity inflow+
¸
All elements
Element controls
The critical path method equations provide the information on the redun-
dancy, and hence constitute part of the model at the project level.
Synthesis over levels 291
Input
Output
Project
Activity
Activity
Activity
Activity
Figure 9.11 Example project composed of activities (not to scale); activity-on-node
version.
From a production viewpoint, the control will be production related.
The arrows between activities (in an activity-on-node context) represent
cumulative activity production.
From a resource usage viewpoint, the control will be activity resources
(number and type). The arrows between activities (in an activity-on-node
context) represent cumulative activity resources.
The inflow to and outflow from a project (at the start and end of
the project respectively) will be cumulative project resources and cumula-
tive project production. For different project level controls (work methods
including activity interaction), different forms of Figure 9.11 will result.
In Figure 9.11, the project inflow–outflow relationship is, for each
resource type,
Project production =Sum of activity productions
=
¸
All activities
¸
All elements
Element productions
=
¸
All activities
¸
All elements
µH
Project resource numbers =Sum of activity resource numbers
=
¸
All activities
¸
All elements
Element resource numbers
=
¸
All activities
¸
All elements
µ
292 Synthesis treatment
The controls at the project level are work method (including sequencing),
and
¸
All activities
¸
All elements
µH and
¸
All activities
¸
All elements
µ, for each resource type; the
behaviour is given by project production and project resource numbers
(quantity), for each resource type.
Resource numbers and type can be measured or expressed in money
terms.
Level planning problems
Introduction
A summary of the various level (constituent, element, activity and project)
problems together with the relevant variables follows.
Constituent level
The constituent level is the basic behaviour level of a resource (person or
piece of equipment).
Control
The control at the constituent level is the resource production rate H for
each resource type.
Output, state
The resource performance is in terms of its production.
Objectives, constraints
Objectives and constraints are resource production or resource production
rate connected for different resource types.
Model
See above.
Problem outcome
Resource production rates and resource production. Commonly, these are
folded up to the next level.
Element level
An element is a single time unit component of an activity.
Control
The controls at the element level are µH and the resource number (quantity)
µ for each resource type.
Synthesis over levels 293
Output, state
Element performance, for each resource type, is indicated by the element
production and the resource number (quantity).
Objectives, constraints
Objectives and constraints will relate to element production, and element
resource number and type (for example, resource availability and applica-
bility).
Model
See above.
Problem outcome
Element production, resource numbers and resource types are commonly
folded up to the next level.
Resource numbers and type can be measured or expressed in money
terms.
Activity level
An activity of duration I is composed of I elements.
Control
The controls at the activity level are
¸
All elements
µH and
¸
All elements
µ for each resource
type and activity.
Output, state
The activity performance, for each resource type, is indicated by activity
production and the activity resource number (quantity).
Objectives, constraints
Objectives and constraints will relate to activity production, and activity
resource number and type.
There may be resource availability and applicability issues. Particular
resource usages may apply, for example those relating to particular work
practices of the work crews and subcontractors.
Model
See above.
Problem outcomes
The resource (or money) usage within activities, and total activity resource
usage and activity production.
294 Synthesis treatment
Resource numbers and type can be measured or expressed in money
terms.
Project level
The project level also includes subproject, sub-subproject and similar levels,
and represents a collection of connected activities.
Control
The controls at the project level are work method (including sequencing),
and
¸
All activities
¸
All elements
µH and
¸
All activities
¸
All elements
µ, for each resource type.
Output, state
Project performance, for each resource type, is indicated by project produc-
tion and project resource number (quantity).
Objectives, constraints
Objectives and constraints will relate to project production, project resource
number and type, and work method.
The total work will have to conform to the (project) scope.
Owner-specified objectives and constraints will carry down to lower
levels.
Model
See above.
Problem outcomes
The work method, and resource (or money) usage within the project, and
total project resource usage and project production.
Resource numbers and type can be measured or expressed in money terms.
General comment
Generally the higher the level, the harder the problem, but the more useful
the information coming out of the problem.
A common approach (subproblem) is to fix the work method, and only
allow choice in resource usage and resource production rates. That is,
only the lower level problems are solved. The possibility of multiple work
methods or networks (involving activity sequencing) is commonly not left
open, but rather assumed known (unless an iterative-analysis approach is
adopted). The synthesis problem of determining a preferred network from a
class of networks is too difficult. It is analogous to Michell frame-structures
in structural engineering (Carmichael, 1981), for which only a few special
results are known.
Synthesis over levels 295
Upon establishing the controls at all levels, then additional outcomes
include:

The timing of the activities. This includes the start and finish times of
activities, allowing a bar chart or generally a program to be drawn

Project milestones and target dates

Critical project activities

Expenditure, budget, cash flow (additionally requiring estimates of
income)

Resource plots

Delivery requirements.
Multilevel optimisation thinking
Introduction
Systems may be decomposed to subsystems with interaction. Likewise, opti-
misation at the system level can be decomposed to optimisations at the
subsystem level together with interaction. Lower level problems could be
expected to be easier to solve, though the coupling between subproblems
introduces iterations into the computations.
The lower level is generally referred to as the infimal or first level, while
the upper level is referred to as the supremal or second level (Figure 9.12).
The upper level coordinates the lower levels, such that the solution to
the original problem is obtained. Extensions to further levels are possi-
ble, although for practical project planning reasons decomposition would
possibly not go below activity level.
Decomposition of the system level problem can be done based on a
physically meaningful decomposition or non-physically meaningful decom-
position of the system, though the former would be more attractive for
planners and could be expected to provide less complicated solutions. For
Supremal
problem
Infimal
problem 1
Infimal
problem 2
Infimal
problem P
. . .
. . .
Figure 9.12 Problem decomposition and interaction.
296 Synthesis treatment
planning problems, the most obvious project decomposition choices would
be according to conventional work breakdown structure thinking:

Phase (chronological); time period phasing

Work type or nature; resource groupings

Work parcels; the way the work is to be carried out

Contracts

Region, location or geographical area

Organisational breakdown

(Existing) cost codes, cost centres, cost headings; project cost and data
summary reports

The nature of the planning network; activity groupings

Function.
But other choices are possible. The decomposition chosen is at the discretion
of the planner.
Phase or stage treatment of planning is also considered in Chapter 8.
There the project duration is broken into subintervals. The subintervals
become the subprojects, and may have some physical meaning, or may be
chosen for some other, perhaps computational, reason.
The solution of a multilevel optimisation problem requires decomposition
and coordination. Decomposition implies taking the project level problem
components – model, objectives and constraints – and breaking these down
to models, objectives and constraints at the subproject level. Interaction
variables are introduced in order to separate the subproblems. Coordina-
tion of the subproblem solutions ensures that the solution to the original
project level problem is obtained. Coordination requires a repeated two-
way iterative transfer between the levels, adjusting the coordination until a
desired solution is obtained.
Generally the solution follows Figure 9.13 or Figure 9.14. A number of
choices are available for selecting the coordination variables.
The networks shown in Figure 9.15 are used in examples in the following.
Decomposition according to the system model
Chapter 8 gives an example staging of the networks shown in Figure 9.15.
There system equations of the following form arise,








x3
x5
x6
x7
x10
x12








=








x1
x3
x3
x3
x6+x7−x3
x10








+








u1+u2
u4
u3+u5
u3+u4
u7+u9+u10+u8−u3
u11+u6+u12








Synthesis over levels 297
Select coordination
variables
Update values of
coordination variables
Fix values of
coordination variables
Solve independent
subproblems
Iteration
Solution
Figure 9.13 Iterative solution to the multilevel optimisation problem.
Supremal unit
Infimal units
Figure 9.14 Iteration between levels.
where xj is the output/state following activity (t, ji, u(t, ji is the control for
activity (t, ji, and where the following has been introduced to simplify the
notation,
u1 =u(1,2i
u2 =u(2,3i
u3 =u(3,4i
u4 =u(3,5i
u5 =u(4,6i
u6 =u(5,11i
u7 =u(6,8i
u8 =u(7,9i
u9 =u(8,9i
u10 =u(9,10i
u11 =u(10,11i
u12 = u(11,12i
u(4,7i and u(5,7i relate to dummy activities and have been omitted.
298 Synthesis treatment
Assemble
options
Apply
finishings
Marketing
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Dummy
Dummy
(a)
Assemble
options
Apply
finishings
Marketing
(b)
Sales
Prelim.
design
Approval,
base cpts
Deliver
base cpts
Assemble
base cpts
Manuf.
base cpts
Approval,
options
Det. dsgn,
options
Det. dsgn
base cpts
Figure 9.15 Example network diagrams. (a) Activity-on-link diagram; (b) Activity-on-node
diagram. (Activity names have been abbreviated because of space limitations –
work items are implied.)
The state represents cumulative resources and cumulative production.
The control represents the chosen resource numbers and production.
Assume a minimum duration (maximum production) objective of
minJ = −
12
¸
t=1
ut
(where only the ut components related to production are considered).
And a constraint related to resource usage,
ut ≤U
t
t =1, 2, . . . , 12
where U
t
, t = 1, 2, . . . , 12, are set resource numbers (and where only the
ut components related to resource usage are considered).
Synthesis over levels 299
Other constraints would generally be present for a meaningful solution.
As well, the objective could be expected to be more involved. However, the
objective and constraint given here are sufficient for the present demonstra-
tion purposes.
The decomposition selected for this example partitions the system equa-
tions – in effect a horizontal line is drawn two-thirds the way down through
the above system equations. This corresponds to the division between stages
1 and 2 in Figure 9.16.
For this particular decomposition, to separate the subproblems, introduce
the interaction variables.
t
1
=x6
t
2
=x7
t
3
=x3
t
4
=x3
Stage 0 Stage 1
Stage 2
Stage 3
u(1,2)
u(2,3) u(3,4) u(4,6) u(6,8) u(8,9)
u(3,5) u(7,9)
u(5,11)
u(9,10) u(10,11)
u(11,12)
x1
x2
x3 x4 x6 x8
x9
x10
x11
x12
x3
x4
x5
x5
x9
x11
u(4,7)
u(5,7)
x7
x7
Stage 0 Stage 1 Stage 2
1
2 3 4
5
6
7
8 9
10
11 12
u(1,2)
u(2,3) u(3,4)
u(3,5)
Stage 3
u(4,6)
u(4,7)
u(5,7)
u(6,8)
u(5,11)
u(7,9)
u(8,9)
u(9,10)
u(10,11)
u(11,12)
(a)
(b)
Figure 9.16 Example network diagrams. (a) Activity-on-link diagram; (b) Activity-on-node
diagram.
300 Synthesis treatment
The objective becomes
minJ = −
12
¸
t=1
ut +3
1
(t
1
−x6i +3
2
(t
2
−x7i +3
3
(t
3
−x3i +3
4
(t
4
−u3i
where 3
t
, t =1, 2, . . . , 4 are Lagrange multipliers.
Subproblem one
minJ
1
=min−
5
¸
t=1
ut +3
1
x6+3
2
x7+3
3
x3+3
4
u3
subject to




x3
x5
x6
x7




=




x1
x3
x3
x3




+




u1+u2
u4
u3+u5
u3+u4




ut ≤U
t
t =1, 2, . . . , 5
Subproblem two
minJ
2
=min−
12
¸
t=6
ut +3
1
t
1
+3
2
t
2
+3
3
t
3
+3
4
t
4
subject to
¸
x10
x12
¸
=
¸
t
1
+t
2
−t
3
x10
¸
+
¸
u7+u9+u10+u8−t
4
u11+u6+u12
¸
ut ≤U
t
t =6, 7, . . . , 12
Solution
One possible solution fixes the values for the Lagrange multipliers 3
t
, t =
1, 2, . . . , 4, and solves the suboptimisation problems for the xs, us and ts.
The values of 3
t
, t = 1, 2, . . . , 4 are then updated such that x6 approaches
t
1
, x7 approaches t
2
, x3 approaches t
3
and u3 approaches t
4
. And the
process repeats.
Other solution approaches are possible, as are other decompositions of
the original problem.
Synthesis over levels 301
Decomposition according to hierarchy
As an example of a problem involving hierarchical decomposition assume,
as before, a minimum duration (maximum resource usage/expenditure)
objective of
minJ = −
12
¸
t=1
ut
and a constraint related to resource usage,
ut ≤U
t
t =1, 2, . . . , 12
where U
t
, t = 1, 2, . . . , 12, are set resource numbers, together with the
example networks of Figures 9.15 and 9.17.
The notation used is the same as in the previous example.
As before, other constraints would generally be present for a meaningful
solution. As well, the objective could be expected to be more involved.
However, the objective and constraint given here are sufficient for the
present demonstration purposes.
1
2 3 4
5
6
7
8 9
10
11 12
u(1,2)
u(2,3) u(3,4)
u(3,5)
u(4,6)
u(4,7)
u(5,7)
u(6,8)
u(5,11)
u(7,9)
u(8,9)
u(9,10)
u(10,11)
u(11,12)
(a)
(b)
u(1,2)
u(2,3) u(3,4) u(4,6) u(6,8) u(8,9)
u(3,5) u(7,9)
u(5,11)
u(9,10) u(10,11)
u(11,12)
x1
x2
x3 x4 x6 x8
x9
x10
x11
x12
x3
x4
x5
x5
x9
x11
u(4,7)
u(5,7)
x7
x7
Figure 9.17 Example network diagrams. (a) Activity-on-link diagram; (b) Activity-on-node
diagram.
302 Synthesis treatment
At the project level, the systemequations can be written using the notation
of Figure 9.17 as,


















x2
x3
x4
x5
x6
x7
x8
x9
x10
x11
x12


















=


















x1 +u(1,2i
x2 +u(2,3i
x3 +u(3,4i
x3 +u(3,5i
x4 +u(4,6i
x4 +u(4,7i +x5+u(5,7i
x6 +u(6,8i
x8 +u(8,9i +x7+u(7,9i
x9 +u(9,10i
x10 +u(10,11i +x5+u(5,11i
x11 +u(11,12i





































x3
x4
x5


















with x1 given. Or a matrix equation in 12 rows could be given if x1 is
included.
Here xj is the output/state following activity (t, ji, and u(t, ji is the control
for activity (t, ji.
In summarised form,
X =M−B
where X, M and B are column vectors of 11 rows, with components xt, mt
and bt respectively. t =2, 3, . . . , 12; or t =1, 2, . . . , 12.
This can be built up from the activity level as shown in the following.
Activity model
(a) At the activity level (Figure 9.18), the system equation for each activ-
ity is
x

j =x

t +u(t, ji
where
x

t the state following node t
x

j the state leading into node j
u(t, ji control for activity (t, ji.
i j
u(i, j)
x
*
i x ′j
Figure 9.18 Activity notation; activity-on-link version.
Synthesis over levels 303
i
x
*
i x ′i
h
.
.
.
.
.
.
j
Figure 9.19 Continuity at a node; activity-on-link version.
Activity interaction
(b) Continuity at a node (Figure 9.19),
x

t =
¸
All preceding
activities(l,ti
x

t
(c) Activity connectivity that accommodates redundant paths through a
network may be expressed by
x

j = −bj
where
x

j correction to the state leading into node j because of redundancies
in the network.
Project model
The project model is obtained by combining (a), (b) and (c) to give
xj =
¸
All preceding
activities
|xt +u(t, ji] −bj
or
X =M−B
304 Synthesis treatment
where
mj =
¸
All preceding
activities
|xt +u(t, ji]
and xj is the state following node j.
Optimisation
To solve the project level optimisation problem as a collection of activity
level subproblems, the activity interaction, the interaction equations (b) and
(c) are adjoined to the original objective,
J

=J +
12
¸
t=1
81
t
|x

t −
¸
all
(l,ti
x

t] +
12
¸
j=1
82
j
|x

j +bj]
where 81 and 82 are Lagrange multipliers.
Consider, for example, the subproblem associated with activity (11,12),
minJ

= −u(11,12i +81
11
x11−81
12
x

12+82
11
x

11
subject to (using the simpler notation of the previous chapter),
x12 =x11+u(11,12i
u(11, 12i ≤U
11,12
Other formulations can be given.
Chapter 10
Selected topic
Project compression
Introduction
The project compression problem is a subproblem of the planning problem.
Typically, the project level control (work method) is held fixed and only
lower level controls are optimised.
A number of approaches have been suggested in the literature for solving
the project compression problem as an optimisation problem, typically
a linear optimisation or linear programming problem. The optimisation
approaches do not appear to have found favour in commercial project
planning programs or with practitioners. Some reasons suggested for
this are:

Most commercial project planning packages appear to be based on
algorithms that reflect the particular characteristics of networks. In this
way the algorithm can be written in general terms and applied to all
networks. A linear programming approach, on the other hand, requires
input peculiar to each network – the network logic of activity prece-
dence is interpreted as constraints in the linear programming approach.
It is conceivable that a transformation could be written in general terms
giving the network logic as constraints.

Knowledge of linear programming and how to formulate problems as
optimisation problems is not widespread. As well, linear programming
packages tend not to be readily available, although some spreadsheets
do include this option.
Nevertheless, it is informative to see the linear programming formulation
to the project compression problem.
The development below follows Carmichael (1996).
306 Synthesis treatment
Cost–duration relationships
A cost–duration relationship or cost function such as Figure 10.1 is typically
assumed.
The equation of the cost–duration relationship is
c
t
= C
t
−s
t
x
t
YC
t
≤ x
t
≤ YN
t
where
x
t
duration of activity t
c
t
direct cost of activity t
s
t
cost slope for activity t
C
t
intercept on the cost axis for activity t
t activity number; t =1, 2, . . . , N
N total number of activities
CN
t
normal cost
CC
t
crash cost
YN
t
normal duration
YC
t
crash duration.
Cost functions other than linear
Many activities could be expected to have convex cost functions or cost
functions that are not linear (Figure 10.2). For convenience, a linear rela-
tionship might be assumed going between the normal and crash points.
Alternatively, a piecewise linear approximation might be assumed.
Activity
direct
cost
Activity
duration
c
i
C
i
y
i
CC
i
CN
i
YC
i
YN
i
Slope =
cost increase
time decrease
CC
i
– CN
i
YN
i
– YC
i
s
i
=
(money/time)
Figure 10.1 Linear cost–duration relationship for activity |; | = 1, 2, . . . , N.
Selected topic 307
Piecewise linear
approximation
Linear approximation
Normal
point
Crash
point
Activity
direct
cost
Activity
duration
c
i
y
i
CC
i
CN
i
YC
i
YN
i
Figure 10.2 Convex cost function for an activity.
The accuracy of any approximation is not an issue here because the
accuracy with which the normal and crash points and all points in between
is known is uncertain.
Where the cost–duration relationship is integer in form, then continuous
approximations might be assumed, or integer (linear) programming used.
Integer-based optimisation techniques do not seem as popular as other
optimisation techniques.
The optimisation problem
Any optimisation problem typically involves three components; an objective
function, a system model and constraints. For the project compression
problem these are as follows.
Objective function
The objective function is usually one of cost and a minimum is sought. The
total (direct) cost is the sum of the activity costs.
minJ =
N
¸
t=1
C
t
−s
t
x
t
or equivalently
maxJ =
N
¸
t=1
s
t
x
t
308 Synthesis treatment
Here minimisation has been replaced by maximisation and a sign rever-
sal. The constants have been removed because these contribute nothing to
establishing the optimum activity durations (but they do contribute to the
total [direct] cost).
Bonus for early completion, penalty for late completion
For a variable project end date, the duration by which the end date is less
than or greater than a nominated duration, T
project
, is
T
project
−LFT
N

where LFT
N
is the latest finish time of the last project activity.
Any bonus, ß, or penalty, I, will be a function of this. Either ß or I
applies but not both.
The objective function becomes
maxJ =
N
¸
t=1
s
t
x
t
+ß−I
ß and I could also be used like penalty functions to force a solution close
to a desired project completion date, in the absence of any project bonuses
or penalties.
Constraints
A number of constraints, restricting the choice of the optimum solution,
apply.
Budget constraints restrict the total amount of money that can be spent
to a fixed amount. That is,
N
¸
t−1
C
t
−s
t
x
t
≤ C
project
where C
project
is the maximum amount of money to be spent on the project.
Where C
project
is not known, it may be left as something to be determined
from the optimisation. That is, this constraint will not apply.
Activity duration ranges give the ranges over which the activity cost–
duration relationships apply and possibly also the workable durations of
activities.
YC
t
≤ x
t
≤ YN
t
t =1, 2, . . . , N
Selected topic 309
Non-negativity constraints, although seemingly obvious, are added to
ensure a sensible solution. That is, all variables are constrained to be non-
negative.
x
t
≥0 t =1, 2, . . . , N
As well, in some formulations, event times may also be variables.
EST
t
≥0
LST
t
≥0
EFT
t
≥0
LFT
t
≥0
TF
t
≥0 t =1, 2, . . . , N
Here,
EST
t
earliest start time of activity t
LST
t
latest start time of activity t
EFT
t
earliest finish time of activity t
LFT
t
latest finish time of activity t
TF
t
total float of activity t
For activity-on-link diagrams, these could be expressed directly as event
times rather than as start or finish times of activities.
Milestone constraints. For various reasons, there may be events that have
to occur before/after some specified time or date.
LFT
t
≤some value for some t
EST
t
≥some value for some t
For example, it may be desirable to set a duration above which the project
duration could not go,
LFT
N
≤some value
For activity-on-link diagrams, these constraints could be expressed as event
time constraints.
Integer value constraints. Generally it could be expected that activity dura-
tions will be required to be integer valued though this need not be the case.
This implies using an integer-based optimisation technique or using
approximations of some form.
310 Synthesis treatment
Subprojects. Where the project consists of subprojects or work packages,
each with their own requirements, similar constraints to all the above will
apply. A reformulation in terms of subprojects or work packages adds
nothing new to the above generalisations; the total number of constraints
is the sum of the subproject constraints together with something relating
the subproject costs to the total (direct) cost.
A subproject is a project. Any formulation for projects is straightfor-
wardly adaptable to subprojects.
System model
The behaviour of the network comes from the precedence relationships as
defined by the network.
That is, the system model is the usual equations that give the earliest start
times, latest start times, earliest finish times, latest finish times and total
floats for the activities, through typically forward pass and backward pass
treatments.
Where overlapping relationships apply, the equations are slightly more
involved than for the non-overlapping relationships case.
Solution
Published approaches to this problem spell out all these constraints, one
by one, even the network logic which is interpreted as multiple constraints.
For a network involving many activities, this can be a lengthy process. An
optimisation package is then used to find a solution.
Example
Consider the problem posed in Deckro et al. (1992). The network is shown
in Figure 10.3. The cost–duration data are shown in Table 10.1.
1
2
3
4 5
6
7
A
B
C
D
E
F
G
H
I
J
K
Figure 10.3 Network diagram for example. Activity-on-link diagram.
Selected topic 311
Table 10.1 Cost–duration data for example
Activity Nodes
(begin, end)
Activity
duration,
lower limit
Activity
duration,
upper limit
Cost function
slope ($/day)
Cost function
intercept ($)
A (1,2) 7 10 5 2131
B (1,3) 6 8 7 3274
C (1,4) 6 13 6 9341
D (2,4) 6 6 – 4070
E (2,7) 20 28 9 2010
F (3,4) 3 5 8 8519
G (3,6) 15 23 11 9708
H (4,5) 3 8 3 434
I (5,6) 4 9 8 7773
J (5,7) 6 10 10 9440
K (6,7) 5 11 6 9542
The project is assumed composed of two subprojects or work packages
comprising activities {A, B, C, D, F¦ and {E, G, H, I, J, K¦.
In the following, x stands for an activity duration, while e stands for
an event time. Both activity durations and event times are variables in this
problem formulation.
Objective function
minJ = (2131−5x
A
i +(3274−7x
B
i +(9341−6x
C
i +(4070i
+(2010−9x
E
i +(8519−8x
F
i +(9708−11x
G
i +(434−3x
H
i
+(7773−8x
I
i +(9440−10x
J
i +(9542−6x
K
i
=66242−5x
A
−7x
B
−6x
C
−9x
E
−8x
F
−11x
G
−3x
H
−8x
I
−10x
J
−6x
K
or
maxJ =5x
A
+7x
B
+6x
C
+9x
E
+8x
F
+11x
G
+3x
H
+8x
I
+10x
J
+6x
K
Constraints
Budget constraints
For an assumed project budget of $65 500,
66242−5x
A
−7x
B
−6x
C
−9x
E
−8x
F
−11x
G
−3x
H
−8x
I
−10x
J
−6x
K
≤65500
312 Synthesis treatment
For separate work package budgets of $27 200 and $38 300 respectively,
(2131−5x
A
i +(3274−7x
B
i +(9341−6x
C
i +(4070i +(8519−8x
F
i
≤27200
(2010−9x
E
i +(9708−11x
G
i +(434−3x
H
i +(7773−8x
I
i
+(9440−10x
J
i +(9542−6x
K
i ≤38300
Due date constraints
For the project to finish by day 31
e
7
≤31
or for the two work packages to finish by days 15 and 31 respectively
e
4
≤15 and e
7
≤31
Activity duration ranges
7 ≤ x
A
≤10
6 ≤ x
B
≤8
6 ≤ x
C
≤13
6 ≤ x
D
≤6 or x
D
=6
20 ≤ x
E
≤28
3 ≤ x
F
≤5
15 ≤ x
G
≤23
3 ≤ x
H
≤8
4 ≤ x
I
≤9
6 ≤ x
J
≤10
5 ≤ x
K
≤11
Bounds on event times
e
1
≤0
e
2
≤0
e
3
≤0
e
4
≤0
Selected topic 313
e
5
≤0
e
6
≤0
e
7
≤0
These non-negative constraints ensure a sensible solution. The non-
negativity constraints on the activity durations can be omitted because they
are taken care of in the constraints relating to activity duration ranges.
System model
The behaviour of the network is given by the network’s sequencing rela-
tionships which are here regarded as additional constraints.
e
2
≥ x
A
+e
1
e
3
≥ x
B
+e
1
e
4
≥ x
C
+e
1
e
4
≥ x
D
+e
2
e
4
≥ x
F
+e
3
e
5
≥ x
H
+e
4
e
6
≥ x
I
+e
5
e
6
≥ x
G
+e
3
e
7
≥ x
E
+e
2
e
7
≥ x
J
+e
5
e
7
≥ x
K
+e
6
Solution
The problem is a linear programming problem in 18 variables (11 activity
durations and 7 event times).
The number of constraints equals 33 and is made up as follows:
Budget constraints 2
Due date constraints 2
Activity duration ranges 11
Event times bounds 7
System model 11
33
A summary of the solution is given in Table 10.2(a, b and c).
314 Synthesis treatment
Table 10.2 Solution summary of example
Activity Nodes
(begin, end)
Activity duration
(days)
Activity cost ($)
A (1,2) 7 2 096
B (1,3) 6 3 232
C (1,4) 13 9 263
D (2,4) 6 4 070
E (2,7) 24 1 794
F (3,4) 5 8 479
G (3,6) 20 9 488
H (4,5) 4 422
I (5,6) 9 7 701
J (5,7) 10 9 340
K (6,7) 5 9 512
65 397
(a)
Node Realisation (days)
1 0
2 7
3 6
4 13
5 31
6 26
7 31
(b)
Completion dura-
tion (days)
Cost ($) Amount below
budget ($)
Work package 1 13 27 140 60
Work package 2 31 38 257 43
Project 31 65 397 103
(c)
A compromise approach
An alternative approach is possible.
Rather than solve the problem as a pure optimisation problem, it is
considered that the strength and efficiency of the forward pass – back-
ward pass algorithm of network analysis, together with the nice features
of optimisation, be employed and the optimisation problem solved in an
iterative analysis fashion.
Selected topic 315
The approach suggested is a comprehensive computer-based method for
the optimal cost–duration problem. The approach contains these features:

Network analysis based on the efficient forward pass – backward pass
algorithm.

An optimisation method based on sequential linearisation which is
capable of solving problems involving large networks with many con-
straints.

The constraints and objective function may be any continuous non-
linear functions involving activity costs and activity durations.
The approach is illustrated by applying it to an example problem.
Using a straight optimisation approach, the requirement of handling
large networks and nonlinear objective functions and constraints, con-
stitutes a dilemma. Also the straight optimisation approach requires a
different formulation for activity-on-link diagrams and activity-on-node
diagrams. A solution to these dilemmas lays in adopting an iterative-
analysis-optimisation approach. Both network types are covered by the
approach. The approach is illustrated schematically in Figure 10.4.
Many of the proposed optimisation approaches are specifically for
either activity-on-link or activity-on-node diagrams. The iterative-analysis-
optimisation approach makes no such assumptions.
All the approaches that use direct optimisation seem to suffer from the
fact that the optimisation problem has to be set up for each case separately.
This is on top of having to set up the network in the first place. The
optimisation problem, for example if linear, requires setting up the left- and
right-hand side coefficients of the constraints.
The iterative-analysis-optimisation approach, on the other hand, only
requires to set up the network which is repeatedly analysed, on proceeding
to the optimum solution.
Within each cycle, the following is carried out:

A network analysis using the latest compressed values, x
0
t
, t =
1, 2, . . . , N, of activity durations. To start the algorithmoff, commonly
activity normal durations might be assumed.

Optimisation is then carried out in order to compress the project further,
that is, reduce some of the activity durations.

These new values, x
0
t
, t =1, 2, . . . , N, of the activity durations are then
input for the network analysis of the following cycle.
The characteristics of the optimisation problem solved within each cycle
are:

Both constraints and objective function, if nonlinear, are linearised
about the latest network configuration, denoted by a superscript ‘0’.
316 Synthesis treatment
Network data
Network analysis
for y
0
Evaluate objective function and
constraints about y
0

(critical activities only)
Linearise objective function and
constraints about latest solution
Linear programming
Updated solution –
Update network data
Convergence
Final solution
Yes
No
Figure 10.4 Iterative-analysis-optimisation scheme of solution.
A nonlinear function, g(xi, may be linearised by using the first order
terms of a Taylor series expansion about a solution x
0
.
g(xi ≡g(x
0
i +(x
1
−x
0
1
i
og(x
0
i
ox
1
+. . . +(x
n
−x
0
n
i
og(x
0
i
ox
n
+higher order terms
where x = |x
1
, x
2
, . . . , x
n
]
T
.
Selected topic 317

Usual activity duration range constraints, budget limit constraints and
non-negativity constraints apply.

The compression is carried out by reducing activity durations one time
unit (day, week, . . . ) by one time unit.
This implies that the linear approximation to the nonlinear objective
function and constraints is reasonable.
But for this restriction on the compression to apply, step constraints
need to be introduced. That is, the activity durations are constrained
from reducing by more than one time unit. These step constraints may
be written as
x
t
≥ x
0
t
−1

Only activities shown to be critical through the network analysis are
included as variables in the optimisation problem.

Milestone constraints are dealt with in the analysis phase, and can be
dealt with as the compression proceeds time unit by time unit.
The algorithm has two features which are not of any real consequence but
may be considered undesirable by some people:

In any one cycle the compression may occur in several critical activities
simultaneously, provided the constraints are not violated. Purists might
suggest that compression should be carried out in order of increasing cost
slopes of activities. This algorithm does not rank the cost slopes unless a
choicebetweencritical activitiesisrequiredtoavoidviolatingaconstraint.

Where parallel critical paths exists, compression occurs on all paths, but
this compression may be by different amounts on each path. What this
translates to is in the analysis on the next cycle, the critical path(s) which
was compressed least becomes critical and the other paths transfer from
being critical to non critical. The optimisation phase then corrects for
this and returns the network to a parallel critical path situation. That is,
there may be a situation of over-compressing or overshooting followed
by a catching up or correction in the next cycle. This over-compressing
can always be checked and allowed for because it gives cost increases
not commensurate with the project duration decreases. Where there are
more than two parallel critical paths, the correction may take more
than one cycle to fully correct.
Example
Consider compressing the duration of a project involving the construction
of a single-span bridge. The relevant network is given in Figure 10.5, and the
relevant cost–duration data and activity explanation are given in Table 10.3.
In addition, an upper cost limit is assumed.
B
28 0
A
5
F
75
G
2
H
7
I
4
J
2
K
2
C
110
D
42
E
28
L
14
M
3
O
2
N
3
Figure 10.5 Single-span bridge construction network (Carmichael, 1989).
Table 10.3 Activity data for example. (Activity names have been abbreviated because of
space limitations – work items are implied)
Activity
Identification
Activity
Description
Normal Crash Cost/
Day
Dur’n Cost Dur’n Cost
(days) (days)
A Move-in 5 40.0 5 40.0 –
B Coffer dam, divert
river flow
28 60.0 24 65.0 1.25
C O¡d

bearings 110 20.0 90 50.0 1.5
D O¡d

precast girders 42 32.0 42 32.0 –
E O¡d

hand rails,
accident barriers
28 6.0 28 6.0 –
F Construct reinforced
concrete abutments
75 100.0 70 130.0 6.0
G Install bearings 2 4.0 2 4.0 –
H Erect precast girders 7 10.0 6 12.0 2.0
I Erect formwork for
deck slab
4 3.0 3 5.5 2.5
J Place deck slab
reinforcement
2 4.0 2 4.0 –
K Pour and finish deck
slab concrete
2 2.0 2 2.0 –
L Cure deck slab
concrete
14 1.0 14 1.0 –
M Lay bitumen surface 3 6.0 3 6.0 –
N Install hand rails,
accident barriers
3 3.0 3 3.0 –
O Clean up 2 4.0 2 4.0 –
Note

O/d – Order and delivery.
Selected topic 319
The steps in the compression analysis are as follows:
Cycle 1
An initial network analysis assuming all normal durations gives the critical
path through the activities O/d bearings (C) – Install bearings (G) – Erect
girders (H) – Erect formwork (I) – Place reinforcement (J) – Pour slab –
Cure concrete (K) – Lay bitumen (L) – Clean up (M). The project duration
is 146 days.
Objective function
minJ =1.5x
C
+2x
H
+2.5x
I
Constraints
Step constraints
x
C
≥109, x
H
≥6, x
I
≥3
Activity duration ranges
90 ≤ x
C
≤110
6 ≤ x
H
≤7
3 ≤ x
I
≤4
Budget limit
295+1.5(YN
C
−x
C
i +2.0(YN
H
−x
H
i +2.5(YN
I
−x
I
i ≤320
Solution
x
C
=109, x
H
=6, x
I
=3
Project duration =143 days
Project direct cost =301
Cycle 2
A reanalysis with updated activity durations for O/d bearings (C), Erect
girders (H) and Erect formwork (I) shows that the critical path remains
unchanged. Activities Erect girders (H) and Erect formwork (I) are at their
crash values and can be removed from consideration.
Objective function
minJ =1.5x
C
320 Synthesis treatment
Constraints
Step constraints
x
C
≥109
Activity duration ranges
90 ≤ x
C
≤110
Budget limit
295+1.5(YN
C
−x
C
i ≤320
Solution
x
C
=108
Project duration =142 days
Project direct cost =302.5
Cycle 3
Parallel critical paths now exist with the path Move-in (A) – Coffer dam
(B) – Construct abutments (F) also critical.
Objective function
minJ =1.5x
C
+1.25x
B
+6x
F
Constraints
Step constraints
x
C
≥107, x
B
≥25, x
F
≥71
Activity duration ranges
90 ≤ x
C
≤110
24 ≤ x
B
≤28
70 ≤ x
F
≤75
Budget limit
295+1.5(YN
C
−x
C
i +1.25(YN
B
−x
B
i +6(YN
F
−x
F
i ≤320
where 320 is the assumed upper cost limit, and 295 is the all normal cost.
(The all crash cost is 329.5.)
Selected topic 321
Solution
x
C
=107, x
B
=25, x
F
=71
Project duration =141 days
Project direct cost =311.25
Cycle 4
In the previous cycle, one of the parallel paths has been compressed by 2
days while the other has only been compressed by 1 day. That is, unneces-
sary expense has been incurred.
The parallel critical path has disappeared and the only remaining critical
path is that with which the solution started.
This cycle rectifies the over compression of the previous cycle.
Solution
x
C
=106
Project duration =140 days
Project direct cost =312.75
Remaining cycles
The remaining cycles, with the budget limit removed, in addition to the
early cycles reveal the progress shown in Table 10.4.
Note that the last cycle overshoots and adds cost without decreasing the
project duration. The fourth, sixth, eighth and tenth cycles are corrections to
over-compressing in the cycle preceding. Over-compressing is recognisable
by there not being a full reduction in project duration commensurate with
Table 10.4 Sequence of project compressions
Cycle Activities compressed Resulting project duration (days)
146
1 C, H, I 143
2 C 142
3 B, F, C 141
4 C 140
5 B, F, C 139
6 C 138
7 B, F, C 137
8 C 136
9 B, F, C 135
10 C 134
11 F, C 133
12 C 133
322 Synthesis treatment
increased cost. Over-compressing then correction is a characteristic of the
approach.
Example – Nonlinear cost functions
A generalisation of the problem solved in Deckro et al. (1992) is considered
here.
The data remain the same as that given earlier but with the linear cost
functions replaced by the following convex cost functions.
Activity Convex cost function
A 2209−23.9y +1.11y
2
B 3370−35.0y +2.00y
2
C 9385−16.9y +0.57y
2
D –
E 2430−45.0y +0.75y
2
F 8564−32.0y +3.00y
2
G 10030−46.6y +0.94y
2
H 444−7.40y +0.40y
2
I 7813−22.6y +1.12y
2
J 9545−38.0y +1.75y
2
K 9579−16.7y +0.67y
2
The linearised versions of these cost functions are obtained from a
Taylor’s series expansion. For example for activity A, this becomes

2209−23.9x
A
+1.11x
2
A

x
0
+

x
A
−x
0
A
¸
−23.9+1.11x
0
A
¸
or
2209+x
A
¸
−23.9+1.11x
0
A
¸
For all activities, these are summarised as:
Activity Linear approximation to cost function
A 2209+y
A

−23.9+1.11y
0
A

B 3370+y
B

−35.0+2.00y
0
B

C 9385+y
C

−16.9+0.57y
0
C

D –
E 2430+y
E

−45.0+0.75y
0
E

F 8564+y
F

−32.0+3.00y
0
F

G 10030+y
G

−46.6+0.94y
0
G

H 444+y
H

−7.40+0.40y
0
H

I 7813+y
I

−22.6+1.12y
0
I

J 9545+y
J

−38.0+1.75y
0
J

K 9579+y
K

−16.7 +0.67y
0
K

Selected topic 323
The solution develops as for the earlier flow diagram, using the linearised
cost functions rather than the original nonlinear cost functions.
References and Bibliography
Butcher, W. S. (1967), Dynamic Programming for Project Cost-Time Curves, Journal
of the Construction Division, ASCE, Vol. 93, No. CO1, March, pp. 59–73.
Carmichael, D. G. (1989), Construction Engineering Networks, Ellis Horwood Ltd
(Wiley), Chichester.
Carmichael, D. G. (1996), An Iterative Optimisation Approach to Project Com-
pression, in Project Management Directions, ed. D. G. Carmichael, B. Hovey,
H. Rijsdijk and D. Baccarini, pp. 149–158, A Special Publication of the Australian
Institute of Project Management, AIPM, Sydney.
Deckro, R. F., Hebert, J. E. and Verdini, W. A. (1992), Project Scheduling with
Work Packages, Omega, Vol. 20, No. 2, pp. 169–182.
Fulkerson, D. R. (1961), A Network Flow Computation for Project Cost Curve,
Management Science, Vol. 7, No. 2, pp. 167– 178.
Goya, S. K. (1975), A Note on ‘A Simple CPM Time-Cost Trade-Off Algorithm’,
Management Science, Vol. 21, No. 6, February.
Kelly, J. E. (1961), Critical-Path Planning and Scheduling; Mathematical Basis,
Operations Research, Vol. 9, pp. 296–320.
Meyer, W. L. and Shaffer, L. R. (1965), Extending CPM to Multiform Project Time-
Cost Curves, Journal of the Construction Division, ASCE, Vol. 91, No. CO1,
May, pp. 45–67.
Panagiotakopoulos, D. (1977), Cost-Time Model for Large CPM Project Net-
works, Journal of the Construction Division, ASCE, Vol. 103, No. CO2, June,
pp. 201–211.
Perera, S. (1980), Linear Programming Solution to Network Compression, Jour-
nal of the Construction Division, ASCE, Vol. 106, No. CO3, September,
pp. 315–326. Discussion, June 1981, pp. 409–410.
Perera. S. (1982), Compression of Overlapping Precedence Networks, Journal of
the Construction Division, ASCE, CO1, March, pp. 1–12.
Prager, W. (1963), A Structural Method of Computing Project Cost Curves, Man-
agement Science, Vol. 9, pp. 394–404.
Robinson, D. R. (1975), A Dynamic Programming Solution to Cost-Time Trade-Off
for CPM, Management Science, Vol. 22, No. 2, October.
Siemens, N. (1971), A Simple CPM Time-Cost Trade Off Algorithm, Management
Science, Vol. 17, No. 6, February.
Index
action 9, 51, 126
activity 40, 114, 169
aggregated 40
artificial 40
continuous 41
critical 48, 51
dummy 40
duration 52, 171, 230
interaction 303
intermittent 41
non-critical 48, 51
production 287
resources 288
status report 198
summary 40
activity-on-arrow diagram 78
actual 19
adages 3
adaptive system 212
analysis 36
iterative 38, 123, 128, 166, 269
arrow diagram 78
as-built 64, 117
as-constructed 64, 117
as-executed 43, 117
as-implemented 64
as-made 117
as-planned 19, 43, 117, 139, 212
backward pass 52, 55
bang-bang control 273, 276
bar chart 41, 74, 97, 112, 138,
172, 251
connected 42, 97, 173
linked 42
reporting 197
baseline 43, 91, 139, 209, 210
behaviour 10
Bellman 13, 277
budget 18, 43, 65, 139
budgeting 43, 65, 112
buffer time period 250
buildability 21, 116
calendar 173
canonical form example 277
case example 11, 15, 16, 22, 123,
133, 144, 145, 160, 162, 164,
182, 184, 189, 206, 207, 226, 254
cash flow 43, 66, 138, 139
change 14, 157, 200
circle diagram 78
communication 137
compression 44, 305
algorithm 48
compromise approach 314
computer package 49, 56, 97
constituent 50
constraint 5, 10, 37, 50, 82, 112,
124, 129, 135, 168, 176, 274,
292, 293, 294, 308
constructability 21, 116
contingency 51, 61, 99
continuous components 255
control 9, 10, 12, 13, 20, 33, 47, 49,
51, 60, 69, 102, 105, 126, 149,
167, 176, 209, 213, 270, 292,
293, 294
automatic 34
bang-bang 273, 276
closed loop 34, 44
error 35, 44, 62
open loop 34, 84, 209
Index 325
optimal 37
singular 273, 276
control systems theory 126
control variables 9, 13, 51, 270
controllability 22, 84, 126
coordination 296
corrective action 13
cost 19, 172, 174, 180, 240
crash 45
direct 46, 63
of financing 67
indirect 46, 63
normal 45
report 193
slope 45
variance 202
cost control system 13
cost–duration relationship 306
cost/schedule control system 205
CPA 51
CPM 51, 56, 91
crashing 45, 93
creativity 21
critical path 51
analysis 42, 51
parallel 48
critical path method 51, 73, 91
criticality index 239
CS
2
205
CSCS 205
cumulative cost 19
plot 112
cumulative delay plot 61, 138, 206
cumulative expenditure plot 66, 111,
112, 175, 194
cumulative income plot 66, 175, 195
cumulative money plot 138
cumulative production 18
plot 43, 57, 74, 97, 138, 205, 249
cumulative resource plot 111, 138
cumulative resource usage 18
data collection 187
decision 9, 51, 126
decomposition 29, 30, 156, 295, 296
hierarchy 301
system model 296
decompression 60
delay 61, 200
reporting 206
deliverable 61, 117
Deming 35
design 5, 6, 8, 136
determinism 128
discrete components 257
disturbance 33, 62, 126, 181, 209
DUR 53
duration 52
crash 45
normal 45
dynamic programming 269, 277
earliest finish time 51
earliest start time 51
earned value 20, 62
reporting 201
EFT 51, 53
element 62
production 286
resources 286
end-product 5, 50, 62, 82, 130, 168
environment 32
EST 51, 53
estimate 63, 171
accuracy 63
event 64, 78
milestone 64, 72
exception reporting 193, 196
expenditure plot 111, 175
fables 3
fast-track 64
feedback 34, 44
FF 52
F/F relationship 52, 53, 84
financial planning 43, 65
float 42, 44, 51, 53, 55, 68
free 52, 53, 55, 68
interfering 52, 68
total 52, 53, 55, 68
forward pass 51, 52, 53
F/S relationship 52, 53, 84
Gantt bar chart 41, 69
gender 25
hammock 40
hierarchical levels 5, 10, 30, 128, 148
hirings and firings 107
IF 52
infimal problem 295
input 9, 32, 33, 51, 126
inverse systems problem 29, 36
investigation 36
326 Index
Kalman 22, 84, 114, 126
key performance indicator 91, 182
KPI 91, 182
ladder representation 89
lag 70
latest finish time 52
latest start time 52
lead time period 53, 69
learning 70, 250
learning curve 70, 226
learning system 212
level 5, 20, 29, 30, 47, 50, 62, 72,
105, 118, 122, 128, 139, 140,
148, 156, 284, 295
activity 151, 171, 286, 293
constituent 151, 172, 285, 292
element 151, 172, 286, 292
planning problems 292
project 150, 169, 289, 294
LFT 52, 53
line of balance 72, 250
linear programming 305
linear project 249
linear–quadratic case 276
linearity 273
link 78
LST 52, 53
LT 53, 69
Markov
process 271
properties 275
mathematical programming 269,
277, 283
measures 107
method study 118
Michell structure 269, 294
milestone 72
event 72
reporting 195
model 37, 124, 284
activity 286, 293, 302
constituent 285, 292
element 286, 293
hierarchical 125, 149, 284
multistage 125, 269
probabilistic 125
project 289, 294, 303
system 272, 275, 277, 302, 310
modelling 29
monitoring 12, 73, 180, 186, 209
frequency 187
Monte Carlo simulation 73, 92,
129, 228
multilevel optimisation 295
multiple activity chart 43, 73
multistage planning 269
Murphy’s Law 15, 50, 64, 74, 138
network 42, 78, 170
activity-on-link 43, 78, 115, 170
activity-on-node 56, 78, 115, 170
analysis 42, 82, 172
directed 78
factual 64, 200
representations 56
time-scaled 97, 114, 138, 172
types 78
node 78
noise 126
non-canonical form example 279
objective 5, 10, 37, 50, 82, 112, 124,
129, 130, 135, 168, 176, 181,
269, 274, 292, 293, 294
function 270, 274, 307
OBS 157
observability 22, 84, 114, 126
optimal control theory 37, 213,
269, 275
optimisation
linear 305
multilevel 295, 304
problem 275, 307
organisational breakdown
structure 157
output 10, 32, 33, 84, 126, 292,
293, 294
overlapping relationships 52, 53,
69, 84
finish-to-finish 52, 64, 84
finish-to-start 52, 64, 84
start-to-finish 52, 84
start-to-start 52, 64, 84
parallel schedule 60, 102, 250, 261
Pareto rule/principle 90, 186
Parkinson’s Law 90
peak requirement 107
people 220
involvement 221
Index 327
performance 10, 34, 212
measure 91, 181
PERT 56, 91, 229
plan 93
planning 10, 35, 93, 136, 180,
222, 270
barriers 223
defined 8
dimensions 122
effort 21
human resource 69
level problems 292
multistage 269
need 5, 7
over levels 148
over total project duration 270
problem components 124
process 122
simplifications 127, 148
Pontryagin’s maximum principle
269, 277
precedence 93
precedence diagram 78
principal of optimality 13
problem-solving process 110
process chart 94
flow process chart 93
outline process chart 94
resource type flow process
chart 94
production 18, 96
rate 96
production/schedule variance 202
program 29, 41, 57, 96, 97, 113,
116, 138
over levels 140
time horizons 141
program evaluation and review
technique 56, 91, 229
progress report 17, 198
project 5, 82, 97, 130, 168
compression 44, 305
fast-track 64
linear 249
production 289
resources 289
stripped-back 29
project close-outs 24
project control 13
project control system 13
project management 98
prophecy 9
pseudorandom number
generator 234
publications 25
quotations 3
redundant link 56
reengineering 118
replanning 12, 35, 99, 129, 139,
179, 180, 224
representation 209, 263
reporting 13, 17, 19, 99, 180,
190, 209
bar chart 197
common troubles 192
cost 193
cumulative delay plot 205
cumulative production plot 205
delay 206
earned value 201
exception 193, 196
factual network 200
incident 193
milestone 193, 195
popular approaches 190
schedule 196
typical report 192
variance 193
reserve 51, 61, 99
resource 8, 9, 99, 171, 174
balancing 60, 100
constrained scheduling 103, 106
histogram 102
levelling 103, 106
multiple types 105, 108
number 100
planning 102
plot 102, 138, 139
production 285
production rate 285
profile 102
smoothing 103, 106
type 100
usage 18, 271
utilisation 107
resource schedule 60, 102, 250, 262
response 36
responsibility matrix/chart 109, 223
risk 110
risk management 110, 129
process 110
328 Index
S curve 66, 111, 138, 139, 194, 202
schedule 18, 41, 57, 112
reporting 196
scheduling 112
scope 112, 129, 134, 136, 169,
212, 213
S/F relationship 52, 53, 84
singular control 273, 276
slack 68
S/S relationship 52, 53, 84
stage 114, 211, 269, 270
modelling 269
state 10, 13, 33, 114, 126, 269, 270,
292, 293, 294, 298
equations 272, 275, 277
space 17, 201, 213
trajectory 13, 17, 273
variables 10, 33, 114, 270
subactivity 85
subproblem 300
subproject 29, 97, 148, 155
subsystem 29
sum of squares 107
supremal problem 295
synthesis 29, 36, 122
optimal 37
over levels 284
problem components 36
system
adaptive 212
learning 212
single level 32
systematic problem solving 38
systems problems 36
systems thinking 29
target 43, 91, 209
task 40, 114
terminology 24
TF 52
time-chainage chart 57, 114,
251, 255
time horizons 139
time management 8, 114
time reserve 61
timetable 116
trajectory 13
trend graph 57
trend plot 57
uniform random numbers 230, 234
uniqueness 26
value management/analysis/engineering
21, 116, 123
variance 17, 117, 212
cost 203
production/schedule 203
variance reduction 244
WBS 117, 152, 157
what if 129
work breakdown 64, 117, 148, 152,
169, 222
work breakdown structure 40,
117, 152
work measurement 118
work method 118
work packages 202
work study 21, 95, 118
Annual subscription packages
We now offer special low-cost bulk subscriptions to
packages of eBooks in certain subject areas. These are
available to libraries or to individuals.
For more information please contact
webmaster.ebooks@tandf.co.uk
We’re continually developing the eBook concept, so
keep up to date by visiting the website.
eBooks – at www.eBookstore.tandf.co.uk
A library at your fingertips!
eBooks are electronic versions of printed books. You can
store them on your PC/laptop or browse them online.
They have advantages for anyone needing rapid access
to a wide variety of published, copyright information.
eBooks can help your research by enabling you to
bookmark chapters, annotate text and use instant searches
to find specific words or phrases. Several eBook files would
fit on even a small laptop or PDA.
NEW: Save money by eSubscribing: cheap, online access
to any eBook for as long as you need it.
www.eBookstore.tandf.co.uk

Project Planning, and Control

Also available from Taylor & Francis Construction Project Management
Peter Fewings and Martyn Jones Taylor & Francis Hb: 0–415–35905–8 Pb: 0–415–35906–6

Construction Safety Management Systems
Steve Rowlinson Spon Press Hb: 0–415–30063–0

Human Resource Management in Construction Projects
Martin Loosemore, Andrew Dainty and Helen Lingard Spon Press Hb: 0–415–26163–5 Pb: 0–415–26164–3

Occupational Health and Safety in Construction Project Management
Helen Lingard and Steve Rowlinson Spon Press Hb: 0–419–26210–5

The Management of Construction Safety and Health
R. J. Coble, T. C. Haupt and J. Hinze Spon Press 90–5809–328–X

Information and ordering details For price availability and ordering visit our website www.tandf.co.uk/ builtenvironment.com Alternatively our books are available from all good bookshops.

Project Planning, and Control

David G. Carmichael

First published 2006 by Taylor & Francis 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Simultaneously published in the USA and Canada by Taylor & Francis 270 Madison Ave, New York, NY 10016, USA Taylor & Francis is an imprint of the Taylor & Francis Group © 2006 David G. Carmichael

This edition published in the Taylor & Francis e-Library, 2006.
“To purchase your own copy of this or any of Taylor & Francis or Routledge’s collection of thousands of eBooks please go to www.eBookstore.tandf.co.uk.”
All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any efforts or omissions that may be made. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging in Publication Data A catalog record for this book has been applied for ISBN 13: 9–78–0–415–34726–6 ISBN 10: 0–415–34726–2 (Print Edition)

Contents

About the author Preface Notation
PART I

vii viii x

Conventional treatment from a systems perspective
1 Introduction Background 3 Replanning 12 Planning effort 21 Book outline 24 2 Systems thinking Introduction 29 Systems–subsystems 29 Single level system 32 Fundamental systems problems 36 3 4 Planning tools and terminology A–Z The planning process Introduction 122 Planning problem components 124 Simplifications to planning 127 Objectives and constraints 129 Scope 134

1 3

29

40 122

vi Contents

Extras feeding into and out of planning 137 Programs 138 5 Planning over levels Introduction 148 Hierarchical modelling 149 Work breakdown 152 Iterative analysis approach 166 Example 168 6 Replanning Introduction 179 Monitoring 186 Popular approaches to reporting 190 Replanning representation 209 7 Selected topics People issues 220 Monte Carlo simulation 228 Linear projects 249
PART II

148

179

220

Synthesis treatment
8 Multistage planning Introduction 269 Stage modelling 269 Optimisation problem 275 Examples 276 9 Synthesis over levels Direct synthesis approach 284 Models 284 Level planning problems 292 Multilevel optimisation thinking 295 10 Selected topic Project compression 305 Index

267 269

284

305

324

Australia. formerly a Graded Arbitrator with The Institute of Arbitrators. In addition there have been numerous smaller consultancies in the structural. He has provided expert reports and expert witness in cases involving structural failures. Major consultancies have included the structural design and analysis of civil and building structures. Australia. the administration and control/replanning of civil engineering projects and contracts. the planning and programming of engineering projects. MEngSc) and The University of Canterbury (PhD) and is a Fellow of The Institution of Engineers.About the author David is a graduate of The University of Sydney (BE. with current strong interests in all phases of project management. teacher and researcher in a wide range of engineering and management fields. He is former Head of the Department of Engineering Construction and Management and currently a Consulting Engineer and Professor of Civil Engineering at The University of New South Wales. construction management and dispute resolution. He has acted as a consultant. construction accidents and safety. a Member of the American Society of Civil Engineers. He is the author and editor of seventeen books and over sixty-five papers in structural and construction engineering and construction and project management. . and contractual and liability matters. construction and building fields. and a trained mediator. and various construction and building-related work.

Critical Path Methods in Construction Practice. M. 1989) was written. service industries to technological endeavours. Woodhead. state space and related terminology of modern control theory is preferred. Central to project management is planning. PERT – program evaluation and review technique. The author was fortunate enough to learn the fundamentals from James Antill (see for example. as is undertaken on all projects on a daily basis. 1965). people are going through the motions of planning but do not understand the fundamentals of their trade. Carmichael. The book assumes that the reader is familiar with elementary planning. Today such packages are universally used. both deterministic (CPM – critical path method) and probabilistic (for example. and Monte . with control systems terminology preferred for its precision and usefulness over conventional planning terminology. And that was at a time when reasonably user-friendly computer packages were available. Ellis Horwood/John Wiley and Sons. 1st ed. Most matters in managing projects. Antill and R. There have been many contributors to project planning. New York. A systems approach is adopted. J. if only as a way of generating bar charts. and this is the basis of this book. Some practitioners were convinced. Project management is now commonplace in most walks of life ranging from the business world. it seemed a losing battle to convince practitioners to use network methods for planning. A main thrust of the book is seen as sending planning thought in the correct direction. involve planning. resources and so on. whether it relates to time. but the majority of planners saw network methods in a negative light. When Construction Engineering Networks (D. As such. Existing treatments on planning jump head first into network analysis. In particular. If used as an introductory textbook. Chichester. The challenge now is to understand what planning actually is. the book treads new ground.Preface It is interesting to watch the development of a discipline over the years. John Wiley and Sons. Planning is a crucial issue in project management and one on which many other aspects develop. supplementary material such as given in Carmichael (1989) would be useful. G. W. costs.

Rigour in the usage of terminology is also stressed. The book rethinks planning practices. . and outdates thinking in existing texts. planners are only after a satisfactory solution. and exercises support the approach. for which the material is applicable. or a solution that they can live with. The author. on-time pressures and pseudo ‘practicality’ arguments. include universities. building and project management. the solutions to planning problems are not unique. As such. One of the overriding reasons for writing this book is to counter the myriad of misconceptions and thinking errors that exist among writers on planning. A planner may also be under time pressures to come up with quick solutions. and do not spend the additional time searching for the optimal solution. The book presents a totally new approach to project planning. Numerous case examples from diverse industries. is guilty of such an uninspired and pedestrian approach (Carmichael. The level of thinking that goes into planning in many cases is very superficial and cookbook in nature. However. and practitioners involved in projects (all industries). amongst many. Planners are unaware of and do not understand the components of the synthesis problem. with all refinements including overlapping relationships. In fact. and relates this to existing treatments. 1989). Acknowledgment The book contains a number of case examples contributed by many people. A more inspired approach is possible. The aims of the book are to understand project planning and to contribute to thinking on project planning. planning can be shown to be a systems synthesis or inverse problem. planners expediently reverse the logic and deflect attention from their inability to come up with best solutions. It contains case examples and exercises and hence could be used as a textbook. Nothing equivalent exists.Preface ix Carlo simulation). to enable the reader to perform more effectively. The book aims to develop the reader’s professional skills and thinking in the planning component of project work. In most cases. justified falsely by the reason of ‘practicality’. The book is aimed at lecturers and students (postgraduate and undergraduate) in project management. professional organisations in engineering (all types). Their contribution is gratefully acknowledged. the book adopts a systems view to provide a rigorous framework. Currently most people use project management texts that have planning (and ‘control’) as one component. and the use of industry-preferred packaged software. and so they never know where they are relative to the optimum. Organisations. to understand project planning procedures and to gain an insight into the associated skills. architecture. To counter this. when in fact planners do not understand the synthetic nature of their job.

k = 0 1 2 N−1 number of periods/stages planned-for state after period/stage k−1 initial state final or terminal state actual state after period/stage k−1 revised planned-for state after period/stage k−1 variance.Notation Continuous time version t tS tF tnow ut xt zt t time project start time project finish time present time control at time t state at time t output at time t disturbance at time t Discrete time/staged version k N xk x0 xN xA k x k var k uk k period/stage counter. difference between planned-for state and actual state after period/stage k−1 control at period/stage k disturbance. anything that prevents x k+1 being obtained exactly Activity information TF FF IF EST EFT Total float Free float Interfering float Earliest start time Earliest finish time .

Notation xi LST LFT LT SOS Latest start time Latest finish time Lead time period Sum of squares .

.

Part I Conventional treatment from a systems perspective .

.

it is said to ‘lag behind’ schedule. success depends on preparation. people also refer in a redundant fashion to ‘future plans’. or if a project is not going well. (Confucius) Action without planning is the cause of every failure. If you fail to plan. Without preparation. Popular adages of planning There are a number of adages and popular quotations involving planning: In all matters. Mediocrity is so easy to achieve. but can you plan any way other than ahead? In a like vein. there will always be failure. there is no point in planning for it. . you plan to fail.Chapter 1 Introduction Background Many people are aware of the humorous paradoxical sign: It also reflects a common saying.

. 4Ps Poor planning = poor performance. and work the plan. where the ant is industrious collecting food over summer while the grasshopper plays. rather than being preceded by a period of worry and depression. The nicest thing about not planning is that failure comes as a complete surprise. and a need for genuine effort to be put into planning. leading to the conclusion that ‘it is best to prepare today for the needs of tomorrow. The first step relates to the distinction between a project and its end-product. only for winter to come and the grasshopper goes hungry. concludes that ‘one good plan that works is better than a many doubtful ones. and clarify some thinking.’ Aesop’s fable of the cat and the fox. Planning also is mentioned in children’s fables. it is first necessary to take a few steps backwards. any road will take you there. To discuss these in a project context. Peter) If you don’t know where you are going. Everybody has a different idea of what planning is. and other publications on planning: • • Everyone says that there is a need for planning. one of Aesop’s fables gives the story of the ant and the grasshopper. 6Ps Prior preparation prevents poor performance – probably. then you have been successful. Real life is what happens when you make other plans. while the fox knows many tricks but cannot decide which to use and gets caught by the dogs. Even if you do nothing. 6Ps Prior preparation and planning prevents poor performance. If you plan to fail and you do. where the cat knows one trick to escape the dogs and uses it. you will end up somewhere else. it should be done on time. Plan the work.4 Conventional treatment from a systems perspective If you don’t know where you are going. Therein lies the source of most of the troubles preventing the advancement of the understanding of planning. (L. For example.’ Two noticeable things emerge from these adages and fables. The stability of a plan tends to vary inversely with its extension.

It is not. service and so on. (The plural terms objectives and constraints are generally used in this book even though the singular forms may be applicable in some situations. They lead to establishing the scope.) The former influences what the end-product will look like and how it functions.) There will be objectives and constraints relating to the end-product. Time Concept Implementation Development Termination Project Idea for end-product End-product exists End of life for end-product Operation.Introduction 5 Project versus end-product A project will commonly come about because of an identified need or want for some product. Broadly on a project. The term ‘objective’ here is used in a correct systems synthesis sense. A project is composed of activities. For example. Figure 1. facility.1). design and planning functions are seen as parallel and fundamental undertakings on projects. Later chapters develop a hierarchy of objectives and constraints – at project. maintenance Figure 1. This distinction sometimes causes people confusion and many people are not aware of the distinction. 2004). An activity is composed of elements. The objectives and constraints are derived from the value system of the project/end-product owner. the exception is where the singular is deliberately intended. Elements are composed of constituents (Figure 1. people sometimes refer to a building as a project. and planning practices.2).3 applies. The latter influences the way of getting to the end-product. activity. The processes that go together to materialise the building are the project. However. and objectives and constraints relating to the project. element and constituent levels. (It is acknowledged that the definition of a project is sufficiently flexible to include the operation and maintenance phase of a product within what is called the project. this is not the issue here. Planning and design Following the distinction between an end-product and the means to the endproduct. The building is the end-product. asset. not as used loosely by lay people (Carmichael. or of the need to make a distinction. This end-product is achieved through a project (Figure 1. .1 Project and end-product timeline.

design texts and design courses do not have to convince readers and practitioners of the need for design.3 Fundamental project undertakings. This indicates the immaturity of project management as a discipline and as practised.2 Hierarchical levels in a project.6 Conventional treatment from a systems perspective Project level Time Activity level Element level Constituent level Resource behaviour Figure 1. there is a section on planning. Both the planning and design problems are shown later to be inverse (synthesis) systems problems. Design Interaction Planning End-product Means to end-product Figure 1. it is accepted as fundamental in order to get to the end-product. and to justify the pages allocated to planning. Why then is not the parallel project task of planning accepted as fundamental. By comparison. design something for them. who has no design education. although popular practice is to solve these in an iterative analysis fashion. and invariably the text authors feel obliged to convince the reader that there is a need for planning and of the benefits of planning. The same holds for courses given on planning. and let the discussion get straight on to the best way to do it? Also of interest to note is that no owner would let a person. It is interesting to note that in nearly every project management text. Owners are very discerning .

or more appropriately the immaturity level of planning compared to the maturity of design. Whether people are involved with projects or not. they are talking about structured planning. the task or project does not proceed. but can soon forget about the cost and duration impact of getting to the end-product. Having breakfast and getting dressed in the morning. Another form of undesirable imbalance can be observed in some project managers and in some project management publications which take the view that activity scheduling is project management. Without planning. The false belief in the need for planning Everyone says that there is a need for planning. as is the question ‘why plan?’ posed in books. The planning might be casual or very detailed. Perhaps when people talk of a need for planning. everyone carries out planning in some form covering the tasks that go to make up their daily lives. and undertake planning for that same ‘discerning’ owner. Planning is carried out (to varying degrees of sophistication) on every project whether people realise it or not. Yet anyone can get away with calling him/herself a planner without any planning credentials. if you have a computer package that does scheduling. it might be well thought out or poorly thought out. the latter is acceptable (it is not unreasonable to expect that effort put into planning improves the chance of project success) but the former implies that no one recognises planning as a fundamental project task. commuting between home and work. driving from A to B. yet it is practised by people with specialist knowledge in the inverse ratio. People have incompatible standards and expectations when it comes to design and planning work. Different planning leads to the task or project being carried out differently. and work tasks. This applies to their home roles. Since planning is a fundamental project task. It may also be that planning is an order of magnitude more difficult than design. a project doesn’t proceed. as opposed to ad hoc planning. talk of need for planning is off the mark. then you have all that is needed to manage a project. But it is more likely that people don’t understand the place of planning.Introduction 7 when it comes to design. rather than talking of the no-planning versus some-planning scenario. Without planning. . social pastimes. are all planned to some degree. and a need for genuine effort to be put into planning. But the real reason appears to be the different maturity levels of design and planning. spending a weekly pay cheque etc. It would be nice to think the reason for this was that the owner has to live with the visual impact and functionality of the end-product. Yet design and planning are both fundamental project tasks on which the project outcome equally depends. it might be over short or long time horizons.

time cannot be managed. Strictly. ‘Through the Looking Glass’. And if lay persons.4). Pareto ) and not important ) activities Resource matters (type. ‘Resources’ are people and equipment. Project management texts give a variety of definitions. neither more nor less’. The parallel matter of design concerns itself with the ‘end-product’ problem.8 Conventional treatment from a systems perspective What is planning? Everybody has a different idea of what planning is. Resource type. in what order and when and with what resources (type. number (quantity) and production are associated with method. ‘The question is’. Planning establishes how and what work will be carried out. the activities involved in the planning process can be treated as another subproject in its own right and planning applied to these activities. on: • • • • Schedule matters (project completion date. differentiation between important (critical. Ch. the actual activities involved in the design process can be treated as a subproject in its own right and planning applied to these activities. Planning outcomes provide information. .) ‘How’ implies ‘method’. and vice versa. This sort of logic can be continued ad infinitum. and order and timing (start. ‘whether you can make words mean so many different things’. for those who like embedded problems. ‘Order’ implies ‘sequence’. and number or quantity) are typically expressed in a common measurement unit of money (Figure 1. Similarly. (Lewis Carroll. for example.’ Humpty Dumpty said in a rather scornful tone. ‘it means just what I choose it to mean. number or quantity. but this is too restrictive. practitioners and texts on ‘strategic planning’ and ‘market planning’ are consulted. for example. and not correct. including outsourcing) Production matters (related to schedule) Money matters (project cost. cash flow ). milestones. duration and finish) of activities. said Alice. and timing requirements. it appears that the word planning is used by most people in the sense of Alice in Wonderland: ‘When I use a word. 6) Some people refer to planning as ‘time management’. and number or quantity) (additionally expressed in a money unit). In broad terms: Planning concerns itself with the ‘means to end-product’ problem. Resources (type. Of course. for those so inclined. See the above adages. (This last sentence refers to the design function.

and number or quantity and hence money) and resource production rates (or equivalent). decision). resources (type. 1975. u t = method t resources t resource production rates t . The project is viewed as a dynamic system. that nothing man strives towards is relevant. It is essential for thinking man to believe in free will. t is time. London. Oscar Wilde observed in Phrases and Philosophies for the Use of the Young that. Each man receives the future dependent upon his earlier actions. Prophecy denies this and declares that all futures are immutable and fixed.4 Some possible relationships between resource usage and money. (Extracted from E. u t has multiple parts (that is. u t . Prophecy Planning is distinguished here from prophecy. throughout the project duration. All tend to be strongly related. The Prophecies of Nostradamus. to believe that his future can be changed by thought and action. Cheetham. as is already preordained by whatever governs his future. Time is waste of money. the answer may lie in Dunne’s concept of alternative variable futures which run parallel to each other. pp. it is a vector quantity of control variables) that contain information on method (including sequence). tS ≤ t ≤ tF .Introduction 9 Money Resource usage Figure 1. planning establishes the value of control (input. be it God or Destiny. action. Corgi. But on the other hand. 14–15) The planning function In systems terms.

activity. Carmichael (1981) considers the end-product problem.2. money (in total or as a function of time) may also be present as a constraint. Frequently a dominating constraint is that of a desired project performance or behaviour (later referred to as state x t ) over the project duration. for example minimum duration. In this last sense. the alternative is to be reactive and not carry out forethought as to what might happen. (Project output.10 Conventional treatment from a systems perspective A hierarchical and staged approach to planning might be adopted. and (fore)thought as to what might happen later in time. or Pareto in some other sense. The selection implies knowledge of work practices. Planning separately on a stage-by-stage or even day-by-day basis is suboptimal. Economy here implies an objective involving money. u t is selected based on whatever objectives and constraints are present. In general. but introduces iterations. For example: • Critical/important work items can be highlighted. Control selections early in a project constrain future updated control selections. relate to the project unless otherwise noted. Objectives and constraints. Planning controls are selected over the full time frame of a project for a globally optimum solution. cumulative money usage or cumulative production or equivalent. That is. the end-product problem is not considered in depth in this book. .) Decoupling the hierarchical levels and stages facilitates the solution. most economical or best in some other way. Determinism would generally also be preferred by planners. output and state are the same.) Flow-on aspects to assist management This book pushes a rational and structured approach to planning as being preferential to an ad hoc approach. (Four main hierarchical levels are considered in this book – project. where mentioned below. The associated objectives enable a choice between alternative solutions to give that solution which is optimal or efficient in resource usage. tS ≤ t ≤ tF . is also an indicator of project performance. x t has multiple parts (that is. z t . Attention can be focused on those activities that are critical in terms of achieving a desired project completion date. planning is said to be proactive. element and constituent – Figure 1. but for the stripped-back project formulations given later. the control is chosen so as to extremise (optimise/minimise/maximise) the objectives while satisfying any constraints present. there are many possible solutions (values of control variables) to the planning problem. such a constraint may eliminate most of the potential for optimisation. Planning well-done gives rise to many flow-on aspects that assist the overall project management. it is a vector quantity of state variables) that contain information on cumulative resource usage.

The branch offices were geographically dispersed and the type of problems differed from country to country. Case example – Desktop computer software One organisation. or to separately go through the risk management process. information necessary to conduct their total affairs. Potential trouble areas and obstacles are identified. Risk management practices are implicitly incorporated. decided to standardise on an operating system and applications used by staff on their desktop computers. Project team members and other project participants can be informed of their duties and when they are to contribute to the project and for how long. delays and uncertainties associated with completion dates can be ascertained. Planning outcomes can provide a basis for coordinating the work of project participants. Flow-on aspects to please the project owner Project owners prefer planning well done because of the improved possibility of a project finishing by a desired date and within a given budget. Allowances and contingencies can be reduced. making planning difficult: • • Additional funds had to be allocated. without needing to mention the term ‘risk management’. Stakeholders can be impressed. for example a computer may cost twice as much in one country as another. taking into account technical uncertainties and equipment availability. together with that at project completion). Coordination between simultaneous projects can be facilitated. In some countries. with branches distributed over a large geographic region and in a number of countries. Predicting the influence of changes and delays is enhanced. Communication between project personnel can be assisted. . A consultant was contracted to provide the redesign and specification. The impact of changes. Uncertainties affecting the project can be reduced. additional funds were required to ensure certain officials approved of the private network infrastructure. One particular personal computer brand was selected as the hardware. A more satisfying work environment for project personnel can be obtained where it can be seen that proper planning has taken place. Owners can be provided with information regarding dates and costs (throughout a project.Introduction 11 • • • • • • • All activities necessary for the project are delineated.

. equipment delivered and staff trained. like planning. An additional time period and cost was involved. In other cases the term is used to mean placing some limit or containment on something such as costs.’ Humpty Dumpty said in a rather scornful tone. is used by most people in the sense of Alice in Wonderland: ‘When I use a word. Even if the equipment arrived on schedule. (Something that is ‘out of control’ is unable to be influenced. in many cases something no more sophisticated than handling data. Delays were common in some countries. Some local vendors caused some logistics problems that delayed the project schedule. Certain specified hardware and software was unavailable in some countries. This required staff to be retrained in order to do the implementation and provide support after implementation. an additional year and a half was required for the project over that originally planned – getting the infrastructure installed.) It appears that the word ‘control’. but really refers to something else. ‘it means just what I choose it to mean. No provision was made for additional parts/spares for replacement.12 Conventional treatment from a systems perspective • • • • • • • • While performing the installation of the equipment. unforeseen problems occurred during the installation of the servers. Equipment had to be sourced elsewhere. an additional time period and cost should have been included in the planning for this project. With all the problems encountered. Replanning Generally the term ‘control’ is used loosely in the literature and by lay persons and practitioners. and lay terms. or keeping this something from growing unacceptably. neither more nor less’. which meant that other sites had to be rescheduled as well. There was a lack of skilled technical personnel. but then not do so. This caused delays. dominance or influence. A vendor would promise equipment to be delivered based on a contract. the term ‘control’ is used in the context of power. Particularly at fault is the usage in some computer packages. where package marketing promotes the feature of ‘control’. In general management texts. there were intermittent power failures that resulted in the installation having to go back to the start of the procedure. Additional staff had to be sourced from outside the organisation to assist in the implementation. In hindsight.

What replanning. ‘whether you can make words mean so many different things’. which was devised for different purposes.’ The words ‘decision’ and ‘control’ may be interchanged. Replanning does what is essentially embodied in Bellman’s principle of optimality. ‘Through the Looking Glass’. Ch. Project performance (the project state x t . As such. updated constraints and possibly updated objectives. most of the discussion on planning applies to replanning. the remaining decisions must constitute an optimal policy with regard to the state resulting from the first decision. the term control is used in a systems sense. resources (or money). meaning the actions. said Alice. compared with planned performance (typically in a report) and. their sloppy usage is inconsistent with the optimum control systems usage preferred in this book. if necessary. 6) In this book. and ‘cost control system’ are not favoured in this book because their lay meanings have a tendency to mislead. The controls bringing about corrections may be in the form of method (including sequence). corrective action is taken to send the project in a desired direction (on a desired-state trajectory). is monitored on an ongoing (usually regular) basis. ‘project control system’. (Lewis Carroll. for the remainder of the project. See Carmichael (1981. The principle states: ‘An optimal policy has the property that whatever the initial state and initial decision.Introduction 13 ‘The question is’. tnow ≤ t ≤ tF . Rather. In essence. u t . The controls are chosen to extremise the objectives while satisfying any constraints present. using updated initial conditions. The popular usage of control as stopping something from growing unacceptably or containing something is incorporated through the use of constraints in the synthesis formulation. Accordingly. input or decisions of the planner (decision maker) in order to bring about a desired project (system) performance. The nature of u t is unchanged – multiple parts (vector quantity) that contain information on method. replanning is no different to planning. and/or resource production rates (or equivalent) input to the project. planning (in the form of replanning) might be . the term replanning is preferred to ‘project control’. resources and resource production rates. 2004). It is planning carried out subsequent to any original planning. when necessary. does is come up with revised values of control variables. shortened time horizon. The common terms ‘project control’. Applying changed or different controls amounts to replanning.

There may be changes in the availability of resources (people and equipment). Work rarely is done exactly as planned. method of implementation and owner needs (revised priorities. conditions might be different. sufficient resource numbers are unavailable when required. unachievable. duration and/or cost estimates may have been too optimistic/pessimistic. activity relationships and activity resource demands and production. it can be said that prior work done constrains future (updated) planning actions. inaccurate or contained mistakes particularly with regard to activity duration estimates. (Project) scope also rarely stays constant from project start to project end. it is rare that any project as originally planned is brought to fruition. The original planning may have been unrealistic. Objectives and constraints may also change. there may be delays due to industrial trouble and bad weather. or perform to expectations/requirements. the use of new or previously unused technology may not have been thought through fully. The ‘planning phase’ of a project occupies the total project duration. necessary preceding activities may be incomplete. Planning is based on certain assumptions. There may be changes in design. There may be coordination problems with other groups. activity sequencing used for planning may have been incorrect. As well. There may be changes in the availability of money. The flexibility in choice of controls decreases with time. A changing project and its environment (represented later by disturbance. Material prices may change. rework ). Generally. variations occur. there may be changes in the availability of materials. For example: • • There are inevitable changes relating to technical specifications. and at project implementation. the design may not have been checked thoroughly. equipment may turn out to be not suited to the task. quality or reliability issues occur. t ) lead to the need for continual monitoring and reporting of project progress. Extreme unexpected events may occur including unimagined geotechnical conditions. Hence the constraints are changing as a project progresses. the scope (of work) changes. There may be technology changes or technical difficulties. subcontractors may be unknown. • • • • • • • • .14 Conventional treatment from a systems perspective referred to as an ongoing or evolutionary process. Monitoring and reporting are necessary precursors to knowing whether replanning is necessary or not. There may be vendors/suppliers or contractors who are unable to meet their original target dates.

and the detailed design could not begin until new dimensions. There was a resulting problem in the location of the sugar store and handling equipment. space or logistics problems. The electrical work comprised the electrical installation for the above work and implementing software for the automatic control of the mechanical plant via programmable controllers and computer screens. The mechanical work comprised the installation of product silos and associated valving and pipework. In the initial planning. regulations and interest rates may change. the return on investment was calculated on sales figures based on potential in the market. that is subjective estimates. These are described below. The total budgeted cost of the project was several million dollars but overran by a third. the plant was required to be of a significantly larger capacity. but also overran by a third. competitors.Introduction 15 • • There may be access. There were uncertainties in the planning that resulted in delays in the implementation stage. the economy. The capital component of the project included civil work comprising of a new building to locate a bulk sugar storage and handling plant and an addition of a mezzanine floor to house machinery in an existing building. later in the project. but there was . The time span for the project was estimated to be about a year. Case example – Bulk concentrated food mix The project involved developing a new plant to manufacture a bulk concentrated food mix for industrial use. The market. When the project was initially budgeted. The delivery periods were also longer for plant and equipment. And according to Murphy’s Law: Anything that can go wrong. The architectural design for the building and the layout had been done based on the original estimates. power requirements and prices for machinery were obtained. This project had both overruns in cost and duration. installing new processing and packaging machinery and installing a new automatic cleaning system. The civil work was managed by a professional project management company and executed by contractors. This affected the plant layout. The mechanical and electrical design and installation was managed by in-house project engineering staff. will go wrong. there was no clear estimate of the required capacity for the new plant. The plant was sized based on the forecast budgeted sales figures plus a growth factor. However.

This resulted in further delays in redesigning the building and obtaining development approval from the local authorities. Some sections of the project were completed well in advance. Case example – Land development project In a land development project. time periods for sequential approvals by government departments of various components of construction. There was no project manager for the entire job. There was also no certainty in the production process because the production technique had to be changed for commercial and contractual reasons. During the commissioning stage. The original planned construction duration was about a year. activities included work in dredging (tidal river). dewatering bunds. while others had significant delays mainly due to long delivery time periods of some plant and modifications to the design. Available boreholes represented samples some 200 m apart. bridge piling and piers. These modifications were done on a trial-and-error basis and consequently added to the increase in the duration and cost of the project. the presence of mud was not recognised. while three mechanical engineers and one electrical engineer undertook the design and the supervision of the plant installation and commissioning. unable to be dredged. a number of modifications had to be done to achieve the required product specifications. Undefined material in dredging The chance of encountering marine clay layers.16 Conventional treatment from a systems perspective inadequate space for the new plant size. The plant was basically of an unproven or novel design that was scaled up from the research and development (R&D) laboratory trials. uncertainties recognised included those associated with: weather and flooding. The person responsible for the project was the project engineering manager. Again. sewerage and stormwater drainage. but he was supervising a number of other projects at the same time. In the initial planning. dry bulk earthworks. and contractor performance relevant to schedule and cost. created an uncertainty in the project duration and cost. rock foreshore protection. There was a project management company that looked after the civil construction. In the initial planning. encountering marine clay layers during dredging and dry excavation. and previous experience from construction in this particular area had indicated that no dry marine mud would be encountered and . these engineers were associated with other projects at the same time.

While the tendered project duration appeared reasonable to the consultant and the owner. tabular reporting of these variances. the duration of the delays in achieving approvals was grossly underestimated. This was particularly relevant to delays from a flood event.) Two state trajectories are shown in Figure 1. As the project progressed. a comparison of actual project performance with planned project performance. Changes in environmental legislation had placed more stringent requirements (for example. One useful report is that which shows together all variances. access was not available until approval had been issued. (For dimensions greater than three. However. state). is processed into progress reports.6 – one corresponding to the as-planned behaviour. additional groundwater testing) since commencing the project. A generic form of collective variance reporting is shown in Figure 1. Although achieving meaningful work in the future stages. and the other corresponding to the actual behaviour. Material was trucked. or pairwise reporting of these variances might be preferred. This influence was handled by arranging the contractor to carry out advanced work in other approved areas for future stages.Introduction 17 was not allowed for. there were no difficulties in this respect.6. . and the estimated activity durations were reflective of expected performance. the owner’s cash flow troubles caused considerable internal management difficulties. as well as absolute project performance. a significant quantity of mud was found in one area. Approval time periods The uncertainty of the approval time period of subactivities by local authorities was recognised in the initial planning.5). In the construction phase. The variances in the states are also shown. in lieu of dredging. Contractor performance The production performance of the contractor was well known for this type of work.6 shows a two-dimensional state space (a space with state variables as axes). that is the difference between as-planned performance and actual performance for all indicators of project performance (that is. For this particular subarea. at an additional cost and an additional duration. In this manner its resources were not left idle (a potential cost to the project without production). which are used as the background to replanning (Figure 1. the space becomes a hyperspace for dimensions greater than three. Figure 1. Information from monitoring. there still remained uncertainty in the activity duration estimates. but the same holds true for where the performance of the project is reflected in more than two states. The response to this was to adopt an alternative method of material removal.

the two states in Figure 1.18 Conventional treatment from a systems perspective Future inf luences estimate Controls Disturbance Plan/Replan Implement Report Absolute and comparative performance Monitor Figure 1. when converted to a money unit.6. As an example. production performance not only reflects actual resource production rates.6 might be cumulative resource usage and cumulative production.5 The replanning process. The variance in cumulative resource usage. but may also reflect delays. The variance in cumulative production will indicate whether a project is ‘ahead/on/behind schedule’. and popular reporting may be more accustomed to schedule and delay reporting rather than reporting as in Figure 1.6 Generic way of representing project variances in a state space. will indicate whether a project is ‘over/under budget’. State 1 Variance in state 2 Planned project state at reporting date Actual As-planned Trajectories Variance in state 1 State 2 State space Actual project state at reporting date Figure 1. in . Popular reporting may also be more accustomed to cost reporting rather than reporting as in Figure 1.6.

6 might then be summarised over time in diagrams such as Figure 1.9. + State 1 variance Date – + State 2 variance Date – Figure 1.7 Progressive plot of state variances. The variances from Figure 1.8 Separate plots of state variances. where individual states are plotted against time.6.Introduction 19 that sense it might be preferred to aggregate the cumulative resource usage states across all resource types and express this as a cumulative cost state. State 1 variance + Date 2 Date 1 – Date 3 Date 4 + State 2 variance – Figure 1. where cumulative resource usage is the state being represented in Figure 1.7 and 1. For example. but says nothing as to whether production is as-planned.9. is shown in Figure 1. less useful than Figure 1. This only gives a part picture of what has happened on the project up to the reporting date.8 to give a historical record of project performance. A way of reporting variances. the figure may indicate that less resource numbers have been used than had been planned. . ahead of as-planned or behind as-planned.

6. Possible replanning controls are many. is no different to the planning problem. Outsourcing part of the work.20 Conventional treatment from a systems perspective State 1 As-planned Trajectories Variance in state 1 Actual Reporting date Time Figure 1. This could be expected to alter the critical paths and the scheduling of activities and costing.. Fast-tracking the remainder of the project. activities done in parallel rather than in series.9 A way of showing project variances less useful than Figure 1. Using new/alternative methods (changed work method). Expediting materials. and resources for the remainder of the project. Example replanning controls The replanning problem. .6 is a preferred form of representation. but still Figure 1.9 to make it more useful (when money and production/schedule are being looked at). and resource production rates (or equivalent). in principle. ranging from altered manning numbers through bonus systems. the use of resources (implying money). Earned value thinking introduces a modification to Figure 1. introducing ‘just-in-time’ or similar inventory practices. for example: Project level controls • • • • • Resources may be redeployed. redistributed. The available controls are the same in both cases. increased or decreased within the project. but commonly centre around method. This implies a new network and new costing. overlapping relationships between activities. replanning may differ in that it becomes an adjustment or change to previously selected controls from the earlier planning. preventative maintenance etc.

bring no additional benefits or savings in total project costs. However. there is a balance between overdoing and underdoing. Unfortunately there is no reliable data on the cost (and amount) of planning. Planning effort As with many matters involving effort and associated cost. technical. 2004). The best solution is that which extremises the objectives while satisfying any constraints present. Still further effort and money put into planning may. Further effort and money put into planning may improve the project outcomes. generally planners are after a satisfactory solution rather than some theoretically optimum solution. though all these are subsumed by a systematic approach to problem solving (Carmichael. Constituent level controls • • Providing incentives Altering resource production rates. Little planning saves money and effort on the planning work but may be wasteful in terms of the total project outcomes (extra cost. which may not be the case.Introduction 21 Activity level controls • • • • Compression (shortening durations) of activities leading to network compression Splitting activities Trading off resource numbers and duration Changing the resource work calendars. late completion. as well as the situation and some creativity. however. Correcting these will also affect the solution. damages ). logic ) affecting the work. Controls selected depend on the planner’s experience and expertise. Some solutions may be better than others. One reason for this is that people cannot agree on what actually constitutes planning and where the boundary between planning and . Through all planning and replanning discussions above. Work study and value management (including constructability/ buildability) studies can be useful here. This list is not totally exhaustive. Element level controls • • Working extra shifts Using extra resource numbers. Replanning may also involve the removal of constraints (physical. it is assumed that estimates and assumptions are correct.

Each project cannot be done multiple times – each time with a changed factor. some may not be present. to a certain extent. but get around the issue to a certain extent by conducting trials on large numbers of people. A ‘control’ project cannot be set up. Performing one project [task] affects other [tasks]. judgement and perhaps logic.) There is only experiential observation or anecdotes of projects to support popular contentions. is not feasible. Experience says that the contribution of particular factors reduces/increases the chance of a project under-performing. Carmichael (2004) argues: A project contains an immense number of influencing factors ([matters] that contribute to bringing about any given result). while all other factors are kept constant. . against which other projects with different factors are compared. while some factors may be more dominant than others. each containing multiple factors that have the ability to influence the projects’ outcomes. To that could be added ‘gut feel’. suffer the same problem. The cost (and amount) will also vary from industry to industry and organisation to organisation. Individual factors cannot be varied one at a time. There are also observability (in the sense of Kalman) troubles. The benefits purported to be due to one factor may be due solely or partly to some other project factors. the contributions of various factors seem reasonable propositions. ( the inverse investigation problem cannot be solved. no two projects are the same. the closest that may be obtained is to observe two similar projects but with different factors. Sloppy usage and understanding of the term ‘planning’ by many people contributes to this. and the outcome of other [tasks]. it might be argued. most of which are interdependent. There is even less data on the relationship between planning effort and project outcomes. although many projects are similar. Conducting trials on large numbers of projects. And on large projects there are large numbers of [tasks]. but this is inconclusive. There is not a one-to-one relationship between any project inputs/factors and project outputs. On any given project some factors may be present. altering any one input/factor affects multiple outputs. none of which is possible to be isolated in order to establish its influence on a project’s outcome. Smaller projects may bear a proportionally higher planning cost than larger projects.22 Conventional treatment from a systems perspective other activities should be drawn. Medical experiments on people. a project (as a system) is almost uncontrollable in systems engineering (Kalman controllability) terms. It is not possible to demonstrate in any objective way how a particular factor contributes to project success. Objectivity is not possible because.

It is not even safe to observe projects and assume that practitioners have homed-in on the optimum by trial and error. Projects are categorised. namely by reasoning that with a certain factor. The shortfall in planning effort also reflected the insufficient knowledge or experience about the job and translated into a poorly constructed job specification that led to a shallow scope (of work) and ultimately unrealistic tender documents. but remain largely subjective. Engineering-based projects evolved from a production department. Often projects were fast-tracked for no apparent reason.Introduction 23 An alternative is to approach the argument from the double negative perspective. medium or low in terms of the factors and the (falsely drawn) conclusion then follows that there is a positive correlation between a factor and ultimate project success. . usually the production manager. Projects. Again this is reasonable. and not any deliberate experimental projects. but can’t be proven. were also done at such a rapid rate that all conventional planning and project staging methodologies were ignored. in some ways. The bottom line result of this was ‘over budget. It is difficult to establish what is the optimum (the balance between the amount of planning and project outcomes) planning effort. in a pseudo-rigorous fashion. Case example – Food production It was common practice within one food production company to undertake projects with much fervour and zeal. then project outcomes will almost certainly be undesirable. promoted the use of fast-tracking in order to circumvent thorough planning. People within the accounting section at the company often argued that thorough planning was costly and could not be justified. little associated planning and a timescale that could not be adhered to. with inappropriate costing. as high. The consequent conclusion is drawn. over schedule’ projects. And this is supported by many case projects that have performed badly with such a factor. The studies give an air of objectivity. Typically they involve examining multiple real projects. This belief. which resulted from operational or competitive necessity. It was felt that thorough planning also led to an increase in project duration and was only a way for some employees to ensure their employment. A number of studies have been reported trying to demonstrate a link between factors and project success. the bane of any project manager.

disposes of or transfers resources including people and closes accounts. Part I takes planning practices (both good and bad) familiar to most planners. This is then viewed as a ‘project’ in itself and planned accordingly. Work in this phase ties together a lot of loose ends. prepares for the transition to take the facility to the operating and maintenance stages. employer. but couches them in the more rational framework offered by systems thinking. Part II offers new theoretical constructions for the planning problem. This elevates the practice of and thinking on planning. files all project data. Personnel problems. Part I relies on the reader having some fundamental exposure to planning. Figure 1. offers considerable insight into what planning is. Some projects engage a new project manager and team specifically to cater for the commissioning and handover.10 shows a schematic outline of the flow of the book. principal. Part II could be omitted in a first reading. Terminology uses Terms used interchangeably by writers include owner. The reason for this is that the systems ideas presented in the book are new to planners. Selected topics in both Parts take a number of interesting planning topics and develop these in a fashion consistent with Parts I and II. and although not intended to be used in everyday planning practice. Planners will benefit by understanding true synthetic approaches.24 Conventional treatment from a systems perspective Project close-outs Special planning commonly goes into the commissioning and handover part of the termination phase of a project. Book outline The book develops the subject in two parts: • • Part I – Conventional treatment from a systems perspective Part II – Synthesis treatment. analyses project performance. Project personnel are aware of the need for thorough planning in the initial stages of a project but some neglect the planning of the project termination. as the project winds down and people are transferred or laid off. . are of particular importance during project close-outs. client. Examples are included throughout the book to reinforce the subject matter. Generally. the term ‘owner’ is the one adopted in this book. developer. proprietor and purchaser.

Synthesis over levels 3. output. ‘him’. response and behaviour Control. .10 Outline of book. The planning process 5. 1981. action. Selected topic 4. Project management is not an exclusive male domain. 306pp. Ellis Horwood Ltd. performance. Influential publications The following books by the author have ideas that were influential in the formation of the planning ideas advanced in this book: Structural Modelling and Optimization. Replanning 7. Selected topics Figure 1. Terminology used interchangeably in this book includes: • • State. Introduction Part II Synthesis treatment 8. Chichester. Systems thinking 9. ‘his’ or ‘man’ when referring to project personnel possibly because the majority of project personnel have historically been male. Planning tools and terminology A–Z 10. Planning over levels 6. (John Wiley and Sons).Introduction 25 Part I Conventional treatment from a systems perspective 1.. ISBN 0 85312 283 0. decision and input. However. such references should be read as nongender specific. Gender comment Commonly references use the masculine ‘he’. Multistage planning 2.

radio and movie programs/guides? How would you like to live in a city without transport timetables? How could you study for a degree without subject. Contracts and International Project Management. 1987. subcontractors and suppliers? 4. Also some people state that they have been associated with projects where no planning was carried out and the projects were successful. Balkema. (John Wiley and Sons Ltd). 435pp. 378pp. Balkema. A. 1989. bar charts.. Exercises 1. Rotterdam. A.. What form of planning might be undertaken by the various project participants – owners. Consider a typical project. Rotterdam. 198pp. contractors. ISBN 0 7458 0706 2. The presence of planning is everywhere.. ISBN 90 5809 325 5. How do you respond to such comments? 6. 5. Disputes and International Projects. time-scaled networks. Consider how people’s routines are affected by the following end-products of planning: • • • • What would you do without television. A. ISBN 0 7458 0212 5. Construction Engineering Networks. 208pp. Chichester.26 Conventional treatment from a systems perspective Engineering Queues in Construction and Mining. television programs. 2002. A. Look around you. ISBN 90 5809 324 7 / 333 6. and cumulative production plots are all examples of ways planning information is conveyed. 2000. ISBN 90 5809 326 3. Lecture timetables. lecture and topic timetables? How would athletes achieve peak performance without a training schedule? 3. training schedules. A. 2004. Rotterdam. (John Wiley and Sons Ltd). consultants. A. Chichester. What is planning to you? 2.. Project Management Framework. The question many people ask is ‘why plan?’ if plans are continuously changing. List benefits of thorough planning. Why do we use one way in one situation and another way in a different situation? Examine some timetables.. transport timetables. Ellis Horwood Ltd. programs etc. What are the disadvantages or drawbacks to thorough planning? . Balkema. and try to improve on the way the planning information is conveyed. Ellis Horwood Ltd. 284pp.

other than given above. Ishikawa. No change in scope is allowed. liquidated damages. and focus people’s attention on the small picture at the expense of the bigger picture? 8. Estimate the cost involved in the planning of a project that you are familiar with. one week and one month. Compare the two amounts estimated. and wages. (In particular. stifle initiative. List possible sources of control that could be used in the following situations in order that the project does not diverge further from the planned state. What suggestions can you make in order to eliminate the causes of these practices? 9. the following planning-related reasons why some projects perform poorly. overheads.) No unethical practices are allowed. with the strongest cause as 1 and the weakest cause as 10. Estimate the cost of delay (avoidable) in the project’s completion date by one day. Base your estimate on component costs such as holding charges (interest on borrowed money). [∗ Planning end-users refer to those who are to use the plan. should planning effort be commensurate with the results required? 10. In what way can a plan limit flexibility in developing new ideas. It is assumed . the specification cannot be altered.] List planning-related reasons. in your opinion. Use a scale from 1 to 10. fishbone diagram or similar might be useful here. Base your estimate on the number of planner-hours that you think are involved.Introduction 27 7. equipment and facility hire. and rank these reasons as well. the so-called ‘doers’. Can the degree of planning observed be justified on an economic basis? In general. Rank. What are the causes of these practices? The use of a cause-effect. as to why some projects perform poorly. – – – – – – – – – – Inadequate period of time and effort spent on planning Lack of ability on the part of the planner Lack of management support for planning Lack of end-user∗ support for the plan Lack of involvement of end-users∗ in the planning Lack of justification of the plan Inadequate or non-existent monitoring Disregarding the plan once developed Monitoring of project performance but no feedback for corrective action End-users∗ not being properly informed of their duties under the plan.

What kind of planning would you undertake? 13. What special planning problems would this create compared to planning being done in a forward sense? . Planning may be done backwards from this point. But how are they related? There is a view put forward on large projects that ‘Extra duration costs extra money. and there is no ‘fat’ available for trimming. Project duration and cost are related. Imagine you are selling your car or house. Plot. • • The project is running behind production/schedule. Amount of planningrelated activities Concept Project start Development Implementation Termination Project end 14. 15. The project is going over budget. Project close-outs commonly have a heavy emphasis on a fixed completion date. Reduced duration saves money. from the point of view of: • • The owner. Broadly. List the conceptual similarities between planning and design. your view of planning-related tasks over the duration of projects. The sale of something can be treated like a project. on the following diagram. Both planning and design are inverse systems problems. 11. This is in contrast to analysis where you determine the output of a system resulting from a given input.’ What is your view? 12. in both problems you are trying to determine the input of a system in order to achieve a desired output.28 Conventional treatment from a systems perspective that the original planning and design for the project have been undertaken in an optimum fashion. The main contractor.

and in particular a synthesis problem. The same type of problem is being solved repeatedly throughout the project. the inverse problem becomes better and better delineated. The interpretation of planning as an inverse systems problem. Systems–subsystems If a project is regarded as a system. uniqueness is established within an optimal control systems formulation. The decomposition may be continued to sub-subsystem (sub-subproject) level and beyond. which would ordinarily be superimposed on this backbone. With the evolution of a project. then just as a system can be broken down into subsystems. are considered separately.Chapter 2 Systems thinking Introduction Framework Much of this chapter is based on Carmichael (2004). both single level and multilevel. Projects in the following have been stripped-back or simplified to their mechanical backbone (Figure 2. with attendant project descriptors of control. so a project can be broken down into subprojects. Thinking on planning is developed in two ways (Figure 2. The conceptual framework offered by control systems theory is regarded as the most suitable for developing planning thinking.1): • • The modelling of projects as control systems. Inverse problems by their nature have non-unique solutions. Issues relating to people behaviour. state and output variables.2). replanning becomes an evolving set of similar inverse problems. (The process can also go in the other direction. from projects to . With project changes and with project evolution.

) Decomposition can be carried out in a number of ways. a ‘program’ being regarded as a collection of projects. for ‘system’ read ‘project’. Resource groupings Work parcels . equipment) Money unit People behavioural skills and issues Stripped-back project Figure 2.3b.30 Conventional treatment from a systems perspective Modelling projects as control systems Planning as an inverse systems problem Multilevel (Optimal) synthesis problem Single level Iterative analysis problem Figure 2.2 Shaded portion showing stripped-back project issues considered in this book. In Figure 2. and a change in the information type.) Commonly. With the example decomposition of Figure 2. in terms of later chapter developments. With hierarchical decomposition of Figure 2.3a represents a two-level hierarchy even though it is called a staged (phased) decomposition. Resources (people. time period phasing Work type or nature.3. ‘program’. project decomposition may be according to: • • • Phase (chronological).3a. the terms ‘higher’ and ‘lower’ are used when referring to levels in such a decomposition. The choice of decomposition will suit the intended project usage.1 Idea development within this book.3. successive decomposition leads to finer detail. the subsystems are all characterised similarly. (Figure 2. for example Figure 2.

(b) Hierarchical decomposition. (a) Staged (phased) decomposition. Activity Element Constituent. location or geographical area Organisational breakdown (Existing) cost codes. Subsystem N System (a) System Interaction among subsystems Subsystems Interaction among sub-subsystems Sub-subsystems Lower level interaction (b) Figure 2. . .Systems thinking 31 • • • • • • Contracts Region.3 Example system decompositions (after Carmichael. Activity groupings Function or generally according to: • • • • • • Project Subproject Sub-subproject. . cost centres. 2004). cost headings The nature of the planning network. work package. Interaction Subsystem 1 Subsystem 2 Interaction . job.

Environment The environment is defined as everything except the system (Figure 2. Subsystems and lower level components also have inputs and outputs.4).5 System as an input–output transformation. and the singular form of the terms is used to refer to both a single entity or plural entities. Input and output contain multiple parts (that is. Subsystems. environment distinction. Single level system To understand systems ideas. share the same properties and descriptors as systems. Multilevel. Constraints. they are vector quantities of input variables and output variables respectively). sub-subsystems.32 Conventional treatment from a systems perspective Subsystems combine through their interaction to give the system at the next higher level. Input System Output Figure 2. Input and output A system may be regarded as an input–output pair or input–output transformation (Figure 2. though sometimes the plural terms ‘inputs’ and ‘outputs’ may be used when the plural is specifically intended. and hence are ‘systems’ in their own right. . The environment affects the system by changes. hierarchical systems (including staging) are collections of single level interacting subsystems. project initial and final (terminal) conditions. and is affected by system changes. Environment System Figure 2. it is convenient to start with a single level system. and disturbance reflect the environment and the environment–system interaction.4 System.5).

(The terms ‘performance’. behaviour or response. it is the project performance or behaviour which is being influenced by the choice of the control. it is a vector quantity of state variables). though sometimes the plural term ‘states’ may be used when the plural is specifically intended. and the singular form of the term is used to refer to both a single entity or plural entities. Disturbance reflects influences (on project performance) that are different to that anticipated when (re)planning (and which prevent as-planned performance being attained). that is there are no observability (in the sense of Kalman) issues. Control contains multiple parts (that is.) Internal system behaviour is usually given in terms of system state. The preferred control is that which extremises the project objectives.6. control Input is equivalent to a decision. decision and action. though sometimes the plural term ‘controls’ may be used when the plural is specifically intended. behaviour or response). action or control. ‘State’ contains multiple parts (that is. state The output describes the external or observable system performance. it is a vector quantity of control variables). Output. output and state are the same. The planner (decision maker) selects the project control in order to get a desired project output (performance. This includes the impact of unforeseen influences Disturbance Input System Output Figure 2.Systems thinking 33 Input. The term ‘control’ is favoured here in place of input. ‘behaviour’ and ‘response’ are used interchangeably in this book. and the singular form of the term is used to refer to both a single entity or plural entities. Disturbance In most systems studies. For the stripped-back projects considered in this book. This corruption is represented by disturbance in Figure 2.6 System with disturbance. Output and state are controlled variables. there is something that prevents the hoped-for output being obtained exactly. . In planning.

It has the better chance of ensuring that the system’s performance goes close to that desired. Closed loop (feedback) control. Automatic control – Technical systems The automatic control of technical (mechanical. Of relevance to planning and replanning are: • • Open loop control. Closed loop control is commonly used in an error control form.7 Comparison of open (a) and closed loop (b) representation. There is a circular transmission of information (Figure 2.34 Conventional treatment from a systems perspective and influences that do not remain constant throughout a project. The control is selected up front.7a). The control is selected and feedback occurs based on the difference between desired system performance and actual system performance.7b). Disturbance Input System Output (a) Disturbance Input System Output Monitor Comparison Feedback (b) Figure 2. chemical ) systems may be done in a number of ways. compared with open loop control where the project performance may or may not go close to that desired. and to keep the system performance as desired. Closed loop (feedback) control attempts to nullify the effects of disturbance on system performance. . and no follow-up monitoring of system performance and corrective control or action is carried out (Figure 2. Information flow is one way. It implies uncertainty. electrical.

Replanning modifications of control systems ideas include: • Controls are not automatically selected. Negative feedback decreases the deviation between desired and actual performance. This gives a project ‘baseline’ or ‘target’ performance. The control is selected with the view that future system performance is going to be affected by future events/influences. The Deming PDCA (plan. do.Systems thinking 35 In error control. there is the possibility of the planned baseline of desired performance (from which errors are measured) changing as the project progresses. Both open loop and closed loop control include anticipation of the effect of future events/influences (much like the approach adopted in risk management). and initial conditions equal to the current project state. Error control ideas are used to guide the choice of which controls need to be adjusted. check. and this in turn can force a change in scope. As well. with the situation of zero deviation being the ultimate target.) There is no feedback if there is no deviation. Planning Planning uses open loop control incorporating estimates of future influences. (Positive feedback increases the deviation. The measurement of project performance is done in most situations with not much accuracy. resource usage and resource production rates. Replanning Monitoring and reporting of project performance gives project status in terms of absolute performance and performance relative to the baseline (errors or variances). act) cycle is no more than a closed loop control system. This introduces (unwanted) deviations from desired performance. but are not used in an automatic closed loop control sense in current replanning practice. Monitoring and reporting follow. for error control purposes. but rather the result of planner intervention. Many future events are not foreseeable or foreseen. feedback may be in a positive or negative sense. Replanning gives an updated project baseline or target performance. making precise control selections not possible. Hence. work method. the relationship between control and performance can be • • . and the cycle continues throughout the project. Replanning uses open loop control incorporating updated estimates of future influences. The planner adjusts the input (control) to a project (which may be subjected to future influences) to give a desired project performance.

time lags within the process interfere with the choice of effective controls.36 Conventional treatment from a systems perspective • • • complicated because of the innumerable things happening on a project at any one time. The output equals the state. There is also no controllability issue. evaluate the control/input. In general the solution of such problems is non-unique. Non-uniqueness in the synthesis and investigation problems might be removed through additionally including some objectives or equivalent. something different is known and it remains to evaluate the unknown. Without the instantaneous use of information coming from the monitoring of the project performance. there is no observability issue. Synthesis – given a model of the system and the output/response. it is assumed that there is no noise in the observation device (though this is not true where people are involved in project monitoring/data collection). Control is applied based on an earlier situation rather than the current situation. the issue is to determine . the fundamental systems problems are: • • • Analysis – given the control/input and a model of the system. whereas the solution of the analysis problem is generally unique. evaluate the model of the system. which is not present in the single objective case. Investigation – given the control/input and the output/response.) Planning and design (and management) are seen to fit within the synthesis categorisation. Both open loop and closed loop control are not used in a real-time sense. Collected project data may be overtaken by new events. Given a certain (desired) response and model. Synthesis requires the a priori specification of a systems model. In the sense of Kalman. Synthesis and investigation are known as inverse problems. evaluate the output/response. (Note that the presence of more than one objective introduces subjectivity. No state estimation is undertaken. that is the relationship between input and output. Fundamental systems problems In terms of Figure 2.5. For the stripped-back projects considered here. Synthesis problem components The synthesis problem is essentially a converse to the analysis problem. In each.

state and output variables. . Generally. The (optimal) synthesis problem is seen to have three main components (Figure 2. objectives (or equivalent terms) resulting from an imposed value statement are implied. Model 2. In addition. optimal synthesis or optimal control selects the control so as to extremise some objectives. Optimal control theory is found extremely useful in elucidating the planning problem. are needed to make the solution unique. These problem components are expressed in terms of control. for example minimum duration or minimum cost. single or multiple objective optimal control problem. Generally one desires to synthesise a system which is optimal in a certain sense. Using state and control variable ideas. Optimal synthesis The optimum problem is within the realm of the well-delineated body of theory and techniques of optimal control systems or optimisation theory. synthesis is equivalent to choosing the project control. Further requirements such as extremisation (optimisation. maximisation or minimisation) of objectives. In simple terms. Constraints. Optimal control theory has applicability across all technology and management. the solution is non-unique.Systems thinking 37 Model Optimal synthesis problem Objective(s) Constraints Figure 2. 1981): 1. Objective(s) 3. the planning problem can be readily exposed and understood. the control that produces this response. a certain response is realisable with many control choices. That is.8) (Carmichael. In this sense the planning problem is a single level or multilevel. See Carmichael (2004) for a rigorous meaning of the term ‘objective’.8 Synthesis problem components. supplementary constraints are also usually present.

Because of the degree of difficulty of solving synthesis problems compared to analysis problems. many synthesis problems are solved in an iterative analysis fashion. Similarly the performance. alternatives Analysing ideas. The iterations arise from the analysis-based mode of attack on the planning problem and are not inherent in planning. established planning practices use iterative analysis.9). That is. the response resulting from the analysis will become more . Feedback occurs between steps in trying to refine the problem (Figure 2. Systematic problem solving goes through the following steps (together with iterative or feedback loops): • • • • • • Defining the problem (including issues and situation) Selecting objectives Generating ideas. alternatives Selecting the best alternatives (evaluation) Action. or equivalently systems engineering. In some cases the analysis problem can even be solved intuitively. It is hoped that by adjusting the initial guesses for the control variables. whereas the solution of the synthesis problem involves more rigorous methodology. behaviour or response is evaluated for each set of these values.38 Conventional treatment from a systems perspective Conversion of synthesis to an iterative analysis problem The solution of the synthesis problem could be expected to be more complicated and more difficult than the solution to the analysis problem. The objectives are evaluated for each set of values of the control variables. Alternatively. The writer’s preferred analysis mode of attack on problems is through the body of knowledge on systematic problem solving. Established planning procedures are iterative in nature. See Carmichael (2004). a systems engineering problem goes through the following steps (together with iterative or feedback loops): • • • • • • Defining the problem (including issues and situation) Selecting objectives Synthesising systems Analysing systems Selecting the best alternatives (evaluation) Action. A systematic approach to the above steps generates clear thinking at each step and lends objectivity to a procedure which would otherwise be considered intuitive.

Beginner planners. although the experience and knowledge of the person attacking the synthesis problem will usually head the iterations in the preferred direction.Systems thinking 39 Problem definition Objectives. experienced planners would be expected to make good first guesses and thereby eliminate the iterations. would be expected to do several iterations. on the other hand. Established. Where this person has synthesised such systems before. The end point of the iterative modification with feedback loops is based upon obtaining satisfactory project performance and satisfactory values for the project objectives. However there is no guarantee of this. constraints Adjust definition. the person’s first guess of values for control variables may be close to ‘optimum’ and no iterations may be required. favourable and also the values of the objectives will improve (decrease or increase as appropriate). constraints Alternatives – guess controls Analysis Adjust controls Is behaviour satisfactory or as required? Yes No Evaluation No Are the values of objectives satisfactory? Yes Synthesis complete Figure 2. nevertheless the synthesis is still being approached via iterative analysis.9 Iterative analysis version of synthesis. objectives. .

a material delivery activity. used to give economy of presentation where detail is not required by the viewer. A dummy activity might be referred to as a logic restraint. where a schedule delay or constraint is required. an administration or a management activity. An artificial activity may be used. An artificial activity which has a finite duration but uses no resources. the level at which detailed duration and resource estimates (and consequently money estimates) are commonly applied. This constraint may be straightforwardly handled by introducing artificial activities into a network. Project start and project end activities may also be optionally introduced in order to give identifiable start and end points to a project. has zero duration and uses no resources. an identifiable piece of work. The term ‘activity’ may refer to a physical activity. certain activities have to be completed by a prescribed date or within a prescribed time period (a duration limitation). Exceptions to this include: • A dummy activity which exists to maintain work method logic in a network. which is a collection or group of activities. The term ‘activity’ is used in this book. An aggregated activity. The terms ‘summary activity’ or ‘hammock’ may also be used. while non-technical disciplines seem to prefer the term ‘task’. Occasionally. or anything that has a finite duration and uses resources (and consequently money). for example. Technical disciplines tend to use the term ‘activity’. • • An activity occurs at the lowest level of a work breakdown structure. where the duration of the artificial activity equals that prescribed time period.Chapter 3 Planning tools and terminology A–Z Activity Also called task. .

The horizontal scale is ideally matched to a calendar allowing for stipulated work hours per day. though usual preference (to facilitate project monitoring) and common practice is to list activities in order of their earliest start times.Planning tools and terminology A–Z 41 Activities may be summarised for end-user purpose according to: • • • • • Description Associated work method. Alternatively. . incorporated materials (including potential suppliers. The activity listing may be in any order. number of shifts per day and so on. Cost account number Assumptions. like activities may be grouped. Bar chart Also called Gantt chart and Gantt bar chart (after Henry L. A bar chart has a timescale horizontally (but no scale vertically). The term ‘bar chart’ is used in this book. A bar chart may be referred to as a program or a schedule. public holidays. But there are other diagrams and charts which are also referred to as programs and schedules. manufacturing and origin) • • • • Activity production rate Duration Money (cost and income) estimate. leading to a stepped appearance from the top left corner to the bottom right corner. Technical disciplines tend to use the term ‘bar chart’. between the start and end times of that activity. with the activities listed vertically. rostered days off (RDO). while non-technical disciplines seem to prefer the term ‘Gantt chart’. weekends. A horizontal bar is drawn adjacent to each activity’s name showing that activity’s scheduled duration. Gantt’s work in the early 1900s). personnel and organisational responsibilities) – Equipment (including facilities. A continuous activity is one that once started continues until complete. A bar chart contains information at both the activity and the project levels. An intermittent activity is one that can be stopped and started with breaks in between. origin and freight) Logic sequence (relationship to other activities) Specification and contractual related matters Resource needs – People (engineering and other services.

the scheduled duration of an activity is shown by a hollow box. In Figure 3. options Marketing Sales Scheduled duration Float Link Figure 3.1.) . The information from a network analysis is readily transferable into bar chart form. indicating the order or logic of work or activity dependence.42 Conventional treatment from a systems perspective and float may also be added. See the associated networks. The practice of drawing lines on a page (without going to the trouble of a critical path analysis) to give the impression of a rationally developed bar chart has little merit. (Activity names have been abbreviated because of space limitations – work items are implied. the influence on following activities of an activity change can be readily established. Activity start 0 10 20 30 40 50 60 Days 80 Preliminary design Approval for base components Detailed design. Activities are preferably shown with float.25(a. Bar charts may be subdivided (horizontal partitions) into subprojects. Such a diagram may be referred to as a connected (or linked) bar chart.1. base components Manufacture base components Deliver base components Assemble base components Assemble options Apply finishings Approval for options Detailed design. Activities are preferably shown connected. as in Figure 3.b).1 Connected bar chart for an example project. float is indicated by a shaded box. Figure 3. a more rational way of obtaining a bar chart is via a network (critical path) analysis because this includes the work method and float information.55. In such a form. It assists both planning and replanning. Figure 3. Although a bar chart can be drawn directly by scribbling horizontal lines on a page. a bar chart is an activity-on-link diagram drawn to a timescale. Activities might be collected together (as also in a subproject) and represented by a single aggregated activity. and time-scaled network.

Cash flow The difference between income (inflow) (payments received) and expenditure (outflow) (payments to suppliers. and activity dependence with other activities. in principle. how long they take. as part of reporting. when they end. As noted. interest on own or borrowed funds ) is the cash flow. as-planned cumulative resource plot). Bar charts also find use in reporting for replanning purposes. budgeting Refer ‘financial planning’. as-planned cash flow). which can be shown in . where actual project performance is included additional to planned performance.1 shows the bar chart plotted for all activities starting at their earliest start times. cost. a connected bar chart is a time-scaled activity-on-link network. Baseline A baseline refers to the as-planned situation. Budget. With connections between activities. production and so on. and delays (as-planned cumulative delay plot). bar chart to network] but this is not the case – the so-called bar chart developed first is in fact a time-scaled [activity-on-link] network. schedule (as-planned program). a large number of bar charts are possible for any given project. the start times of the non-critical activities can be varied because of their float. The connections between activities reflect the logic of the network diagram. The baseline may be with respect to resource usage (as-planned resource plot. employees. It may be referred to as a target. money usage (budget. bar charts and multiple activity charts contain essentially the same information. the amount of float.Planning tools and terminology A–Z 43 and finish times and floats are calculated in the network analysis. trade contractors. Hence. a bar chart gives information on when activities begin. Actual or as-executed progress and performance is compared with the various baselines for resource types. Cumulative production plots. but usually is more readable. production (asplanned cumulative production plot). a bar chart becomes the same as a time-based cumulative production plot. In summary. the bar chart contains the same information as a network. (Some computer packages might appear to go in the reverse direction [that is. This flexibility is used by planners to optimise the usage of resources.) Figure 3. When ‘integrated’. However.

decreasing the project duration may give a lower total (direct plus indirect) cost solution. and all other controls are held fixed. the control is selected based on the difference between desired system behaviour and actual system behaviour. it may be desirable to complete a project early in order to. but introduces more float. A similar situation arises should there be bonuses for early completion. Where there may be large time varying indirect costs. Compression Project compression refers to the reduction or shortening of a project’s duration. See ‘estimates’ for a distinction between direct and indirect costs. industrial disputes. As part of an iterative analysis attack on planning. part way through a project. reuse the equipment or workers/personnel on another project. It is commonly used in an error control form. equipment or worker/personnel shortages may develop during a project thereby delaying the project completion time. Closed loop control Closed loop control may be referred to as feedback control. and might be considered as a possible way of utilising resources more effectively. To assist resource and money allocation on a project or over several projects.44 Conventional treatment from a systems perspective a cash flow diagram or a table. unforeseen workplace conditions ) may occur during project execution. When. Compression may come at a cost penalty. • • • A project’s duration may be reduced by reducing or shortening the durations of critical activities (activity compression). is excessive. A positive cash flow indicates that more money has been received than has been paid out. The project compression problem is a subproblem of the total planning problem. Material. where emphasis is placed on the different level controls that can lead to activity and project duration shortening. based on normal (least cost) durations. say. project compression may be considered: • • Where the calculated project duration. Delays due to other causes (such as those due to weather. the project is running behind schedule. Reducing the durations of non-critical activities does not reduce a project’s duration. In order that the overall project completion time meets some required completion time or deadline. . a negative cash flow indicates the opposite situation. perhaps where penalties or damages for completion time overrun start to accrue.

however. Cost–duration data points represent the minimum costs to perform the activity in the given durations. attention focuses on what are termed normal durations and costs. and this may not be available. Activities with the lower cost slopes are chosen first for compressing because these lead to lower increases in total cost. where it is only possible to undertake an activity in a finite number of ways. . Durations cannot go below activity crash durations – they are the absolute minimum activity durations obtained by employing maximum resource numbers (Figure 3. A theoretical construct called a cost slope can be used to rank activities and establish an order for compressing. The order of compression is selective. Instead. and crash durations and costs. In such cases. Figure 3.Planning tools and terminology A–Z 45 In the compression of activities. As another example. estimating.2 represents the conceptual form of activity cost–duration data. consider a delivery activity. Activity normal durations are those of least direct cost and are usually the durations for which the initial planning calculations are done.3. manufacturers’ or contractors’ information may be limited to only two or three cost–duration data points as shown in Figure 3. or by Further crashing of activity Crash cost Direct cost Cost slope = increase in cost/decrease in duration Normal cost Uneconomical drag out Crash duration Normal duration Activity duration Figure 3. The use of such an approach in practice will depend on having adequate estimating data. but high cost. Generally.2). the item or material might be delivered by air within a short time period. for example by using a team of three workers or a team of five workers. A cost slope may be defined as the relative increase in cost per unit time saved undertaking an activity. The number of possible ways of undertaking an activity may also reduce the data to a few (finite number of) points. Crashing consequently refers to taking compression to the limit. cost–duration data points referring to situations in between these defined points are meaningless. the continuous range of cost–duration information shown in the curve is not available to planners.2 Compression terms.

and the accuracy of this data. that is the cost–duration data is confined to one point and there is only one way of undertaking that activity. time varying indirect costs are less for shorter duration projects. A special form of the discrete case is when the normal and crash points are the same. . If there are large time varying indirect costs. There are many ways that activities can be selected. Project compression may be viable because. A variant of the cost–duration diagram of Figure 3. decreasing the project duration may give a lower total (direct plus indirect) cost solution. but here the choice of which critical activities to compress becomes almost arbitrary.4 indicates the idea where there are two methods available. The interval between these two points is not defined.3 occurs when there are two or more distinct ways of carrying out an activity. These two modes of delivery define two cost–duration points. A possibly more serious issue is the (un)availability of multiple cost–duration data for any activity. sea at less cost but longer time period. A number of writers raise the issue of the validity of assumptions relating to continuity and linearity of cost–duration curves. Compression without activity cost data can also be carried out. However.5). the treatment of this discrete case is similar to that in which a continuum of possible cost–duration data points exists (continuous case). Figure 3. There is consequently a trade-off between increased direct cost and reduced indirect cost (Figure 3.3 Available activity cost–duration data. Any assumption relating to a continuous cost–duration trade-off is not relevant. and hence two distinct cost slopes depending on the method used.46 Conventional treatment from a systems perspective o Available data Crash cost Direct cost o Changing cost slopes o Normal cost o Crash duration Normal duration Activity duration Figure 3. even though the direct costs of activities increase and hence the direct cost of the project increases as the project duration is shortened.

in the compression problem. reward)/stick approach. . Element level • • Use additional resource numbers (quantity) Work overtime or multiple shifts. For control selection.4 Example cost–duration data for two work methods. Direct costs + Indirect costs Project cost 'Optimum cost' Direct costs Time varying indirect costs Time Original completion date Figure 3.Planning tools and terminology A–Z 47 Method A Direct cost Method B Activity duration Figure 3. general ways of shortening a project that fit within the above discussion include: Constituent level • • Change the resource production (constitutive) relationship Use a carrot (incentive.5 Schematic of direct cost and indirect cost trade-off involved in project compression.

Project level (changing method) • • Run concurrent/parallel activities Use overlapping relationships. The solution steps to the project compression subproblem are no different to any project planning problem. There is also the possibility that in order to compress one activity it may be necessary to introduce a new piece of equipment to the project. again. Some oddities that may occur in connection with network compression are as follows. Often compressing of one activity which leads to overtime being worked on that activity may lead to other activities also being carried out on overtime purely to maintain industrial harmony in the workplace even though the project may not require these other activities to be carried out in other than normal durations. these other activities may not require speeding up. Compression algorithm Adopting an iterative analysis approach to planning. the process is complete. • • When no further compression is possible. to manually compress a network. then the compression calculations are complete. The three most notable are outlined below: • Compressing of critical activities may proceed to their respective crash durations and no lower. This piece of equipment may then be used on other activities to gain maximum use out of it although. then any compressing of activities has to be carried out simultaneously and by the same time interval on all critical paths. it is possible by examining a network and the associated cost–duration relationships for the activities. If the status of other activities has not been affected during this process. Compressing on less than the full number of critical paths will not reduce the total project duration.) Compressing of critical activities may proceed until a non-critical activity has its float reduced to zero and hence itself becomes critical (and hence itself a candidate for compressing).48 Conventional treatment from a systems perspective Activity level • Trade-off additional resource numbers (quantity) versus activity duration. or to write a computer program that mimics the manual procedures. Where multiple (parallel) critical paths exist in a network. . The compression calculations may pass through several phases. Coarse steps in the process follow Figure 3. (Similar comments apply to the simultaneous compressing of multiple critical paths.6.

Computer packages There are many computer packages of varying user-friendliness available on the market to assist project planning. and involve editing directly on the monitor and almost instantaneous results for the user. They treat planning subproblems such as resource smoothing. The early packages operating on mainframe computers required punched card input and batch processing with slow turnaround time period between computer runs. but this is ‘control’ in a lay person’s loose meaning of the word. .6 Steps in project compression. They perform the analysis part of the iterative analysis mode of attack on planning. Computer packages based on the critical path method of analysing networks have been around for several decades now. (However. They may advertise ‘control’ capabilities. many established planners still prefer to draft the project network by hand on paper before transferring it to the monitor. resource constrained scheduling and project compression. This was replaced by mini computers and paper tape input. The latest computer packages exploit the graphics capabilities of personal computers. taking into account parallel paths Select activity(ies) with lowest cost increase and which can still be compressed Compress by 1 time unit Figure 3.) The latest packages are also considerably more sophisticated in the number of options they allow and their computer input and computer output capabilities. at best they only provide information on which replanning is based. but in replanning.Planning tools and terminology A–Z 49 (Re)analysis of network Starting solution or all normal solution Identify critical activities Identify non-critical activities Rank in order of increasing cost slope.

Journals of that era published listings of simple computer programs. Refer Carmichael (2004). referring to resource behaviour. Organisations also supplement these planning packages with word processing. Constraints may be stated at all project levels. As with objectives. Constraints restrict the range of control choices possible. the choice of package that an organisation uses is dictated by the package requested by the client. Some packages became well known. Planners need to be careful not to fall into the trap of accepting without questioning anything that comes out of a computer analysis. the packages tend to have common bases. leaving the routine analysis calculations to the computer package but not the decision making (control selection). Today. garbage out) and of course Murphy’s Law (extended) has something to say on the matter: To err is human. the exception is where the singular is deliberately intended. Few organisations write their own software any more. there are many packages in the marketplace. many people confuse a project constraint with an end-product constraint. generalisations of organisations’ in-house packages. All projects have genuine constraints such as funding. and so some adaptability of thinking between packages is required. but all of that era seem to have disappeared and have been replaced or upgraded with more user-friendly packages. . varying in their price and user-options. it was not uncommon for an organisation to have a computer program especially written for scheduling and reporting. in line with the more user-friendly graphical interfaces of modern computers. There arose a number of commercial packages from this background. All programs were based on the critical path method. Most use off-the-shelf packages and adjust their internal procedures to the limitations this imposes. Frequently. Constituent The basic level in a project hierarchy. Both activity-on-node and activity-on-link formats are catered for. spreadsheet and database packages. Fortunately. but to really foul things up requires a computer.50 Conventional treatment from a systems perspective Several decades ago. Constraints The plural terms ‘objectives’ and ‘constraints’ are generally used in this book even though the singular forms may be applicable in some situations. environmental (natural) and political constraints. There is the acronym GIGO (garbage in. A suggested approach is to always interact with any computer package. in an attempt to tap a larger market and get a return on the invested resources writing the computer programs. End-product constraints may influence project constraints.

decision and action. though an 80:20 distribution of non-critical activities to critical activities rarely occurs. The terms ‘critical’ and ‘critical path’ are popularly used terms in the project management field. Critical activity. namely: Earliest start time (EST) Earliest finish time (EFT) . The planner selects the project control in order to get a desired project output (response. The first stage is a forward pass through the network in which earliest times (or dates) are calculated for each activity. If the start or finish of a critical activity is delayed. The term ‘critical path analysis’ (CPA) is also used. The analysis relies on a single duration estimate for each activity. It is possible to have multiple critical paths in a network. CPM The critical path method (CPM) is based on a deterministic network analysis. The preferred control is that which extremises the project objectives. A non-critical activity is one with float.Planning tools and terminology A–Z 51 Contingency Refer ‘reserve’. often used incorrectly. Network analysis gives definiteness to these ideas. performance. amongst the collection of all activities. The critical path is the set of network-connected critical activities from project start to project end determining the project duration. it is a vector quantity of control variables) that contain information on method (including sequence). The network analysis takes advantage of the special directed characteristics of the network. Control Planning establishes the value of control throughout the project duration. It is recommended that most attention be paid to the critical activities during project implementation in order that the project does not run over schedule. A critical activity is akin to a Pareto item. Critical path method. Control has multiple parts (that is. and broadly is taken to include all things connected with such an analysis. behaviour). In particular. resources (and hence money) and resource production rates (or equivalent). the project duration will be extended. it is the longest duration path through a network. the analysis is done in essentially three stages. The term ‘control’ is favoured in this book over input. critical path A critical activity is usually defined as one with no float.

finish-to-finish (F/F). start-to-finish (S/F) and finish-to-start (F/S) – all with finite lead time periods) apply. Forward pass Initial conditions for the first activity(ies): EST = 0 EFT = EST + duration For all subsequent activities: EST = maximum EFT of preceding activities EFT = EST + duration Backward pass Terminal conditions for the last activity(ies): LFT = EFT LST = LFT − duration For all previous activities: LFT = minimum LST of following activities LST = LFT − duration . F/S relationship calculations Mathematically.52 Conventional treatment from a systems perspective The second stage is a backward pass through the network in which latest times (or dates) are calculated for each activity. Similar calculations are performed where other relationships (that is. the calculations are as follows for activities with finishto-start (F/S) relationships (and zero lead time period) between them. namely: Latest start time (LST) Latest finish time (LFT) Thereafter (the third stage) the total float (TF). start-to-start (S/S). free float (FF) and interfering float (IF) may be calculated.

Planning tools and terminology A–Z 53 Float For all activities: TF = LFT − EFT = LST − EST FF = EST of following activity − EFT of activity itself IF = TF − FF Setting the earliest start time of the first activity(ies) to zero is optional. setting the latest finish time equal to the earliest finish time means that the so-called critical activities will have zero total float. That is. activity LFT and LST information is obtained. for example. For the last activity(ies). activity EST and EFT information is obtained and then on a backward pass. The project completion date is the EFT or LFT of the last activity(ies). A full network analysis requires all network interconnections to be considered. The notation of Figure 3. on a forward pass. Overlapping relationships The above calculations can be generalised for all overlapping relationship types. The floats and the critical path may then be determined. Terminal conditions other than the one mentioned above can be prescribed and this would be the case. The relevant expressions are as follows for a network starting at activity 0 and finishing at activity N : Forward pass calculations EST0 = 0 EFT0 = EST0 + DUR0 ⎧ ⎪EFTi + LT ⎪ ⎪ ⎨EST + LT i i ESTj = max all i j ⎪EFTi + LTj − DURj ⎪ ⎪ ⎩ ESTi + LTi + LTj − DURj F/S S/S F/F S/F . The calculations on networks involving overlapping relationships proceed in a similar fashion to the calculations of elementary network analysis. Other values can be prescribed. where there was a desired finish date for the project.7 applies.

(a) Finish-to-start relationship. or overlapping dependence relationship choose the maximum ESTj value because the longest path through the network is sought. Where ESTj turns out to be negative.7 Lead time periods. i. two activities only.54 Conventional treatment from a systems perspective The above implies that for more than one preceding activity. (d) Startto-finish relationship. EFTj = ESTj + DURj EFTi (LFTi ) Activity j Activity i ESTj (LSTj ) LT (a) Activity i Activity j ESTi (LSTi ) ESTj (LSTj ) LTi (b) LTj Activity i EFTi (LFTi ) Activity j (c) ESTi (LSTi ) LTj EFTj (LFTj ) Activity i LTi Activity j (d) EFTj (LFTj ) Figure 3. (c) Finish-to-finish relationship. . (b) Start-to-start relationship. upgrade this value to 0.

and left-hand sides of the earlier expressions for earliest times. Note that where the largest activity LFT is not the network final activity. Free float. LSTi = LFTi − DURi Floats Total float. j. TFN will equal the difference between these two values and the critical path through the network will be defined by activities having total floats equal to this number. . Where the latest finish time is not set equal to the earliest finish time for the last activity. choose the minimum FFi value. In such a case the activities with total float equal to zero will define the critical path. j.Planning tools and terminology A–Z 55 Backward pass calculations LFTN = EFTN LSTN = LFTN − DURN ⎧ ⎪LSTj − LT ⎪ ⎪ ⎨LST − LT + DUR j i i LFTi = min all i j ⎪LFTj − LTj ⎪ ⎪ ⎩ LFTj − LTi − LTj + DURi F/S S/S F/F S/F The above implies that for more than one following activity. or overlapping dependence relationship. These activities can be allocated float in a post-analysis check. TFi = LFTi − EFTi = LSTi − ESTi TFN will equal zero if the latest finish time is set equal to the earliest finish time for the last activity N . ⎧ ⎪ESTj − ⎪ ⎪ ⎨EST − j FFi = min all i j ⎪ESTj − ⎪ ⎪ ⎩ ESTj − EFTi + LT ESTi + LTi EFTi + LTj − DURj ESTi + LTi + LTj − DURj F/S S/S F/F S/F The above implies that for more than one following activity. These expressions represent the difference between the right. the activities at the end of the network (following the activity with the largest LFT) will be calculated to have a zero total float yet they are not critical activities. or overlapping dependence relationship choose the minimum LFTi value.

It has been humorously commented that: The critical path method is a management technique for losing one’s shirt under complete confidence. all have CPM as their basis. Each person.8) is determined by the path A–B–C. network analysis is ‘project management’. The reasoning goes that if you have a computer package. Representation Personal preference dictates the way that the activity start and finish times and floats are displayed on the network diagram. PERT is a probabilistic analysis method and appears not to be used by practitioners (even though the term PERT is used.8 Redundant link example.9 shows a number of examples. This is not a realistic appraisal of project management. To check for redundancy the earliest start time of C has to be computed along the path A–B–C and along the link A–C. redundant links occur with triangular logic patterns as shown in Figure 3. . A B C Figure 3. The numerous commercial computer packages available to assist project planning. then you can set yourself up in business as a ‘project manager’. Redundancy will similarly occur in networks with overlapping relationships provided the earliest start time of C (Figure 3. Figure 3. but wrongly). To many people. redundancy will not occur if the earliest start date is determined by the link A–C. If the value on the link A–C is greater then remove the redundant link A–C. Hence triangular patterns may or may not be permissible in networks with overlapping relationships.8. Occasionally people (incorrectly) use the term ‘PERT’ (program evaluation and review technique) when talking about CPM. Computer packages may allow customising the display. text and computer package has its own preference.56 Conventional treatment from a systems perspective A note For activity-on-node networks with non-overlapping relationships.

For a bar chart such as Figure 3. are: • • • The axis labelling may be interchanged. Some variants on Figure 3. ‘integrating’ the bar chart) and changing the vertical axis gives the cumulative production plot of Figure 3.11).13a.10 can be either an activity level or project level representation. and is used in the line of balance technique.Planning tools and terminology A–Z 57 EST EFT LST LFT TF Activity name Duration (a) Activity name Duration (b) EST EFT LST LFT TF Activity name Duration EST LST EFT LFT TF FF Activity name Duration EST (c) TF EFT Figure 3.13b. Figure 3.12).9 Example drafting styles. trend graph or time-chainage chart. joining the bars (that is. The line plot may be given a finite width where work occupies durations of significance (Figure 3.10.10). at the activity level. The diagram may be drawn down the page (Figure 3. the slope of the plot represents the rate of production (Figure 3. with the independent variable on the vertical axis (Figure 3. in some situations may alternatively be called a trend plot.11). . Cumulative production plot A cumulative production plot shows an activity’s or project’s production up to any time or distance (as the independent variable). A cumulative production plot may be referred to as a program or a schedule.

material usage.11 Example cumulative production plot (time-chainage chart) for road construction. (Activity names have been abbreviated because of space limitations – work items are implied.) Planned rate of progress/ production Learning period Time or distance Figure 3. surface Mar Culvert Apr Seal. . . distance. fencing Figure 3.Cumulative production (tasks. .10 Example cumulative production plot.) . Chainage Cut Finished road line Original ground line Jan Cut Fill Culvert Clear Subbase Feb Subbase Fill Furniture Seal. surface Furniture May Landscaping.

(Activity names have been abbreviated because of space limitations – work items are implied.ber tere r Plas 5 4 Units Carpe nter Plum 1 Working days Figure 3.) Fin is hes 4 ns Carpe ter 3 nter . (a) Bar chart.) R itio Int ern al p art Rou and gh-in s wet ervic area es s Renderer 2 Electr ician Pai n P/R 9 8 7 Floor/storey 6 5 3 2 1 M G Days (a) Figure 3. (b) Cumulative version.12 Example project with sequential/repetitive work.13 Example construction program for internal work for a building. (Activity names have been abbreviated because of space limitations – work items are implied.

) Associated with every cumulative production plot. see ‘resource balancing’. It is the additional vertical scale that makes the plot attractive. The computational path to the cumulative production plot should be no different to that to a conventional bar chart. and all other controls are held fixed. In cumulative production plot terminology. compared to a bar chart. while a parallel schedule is one determined by production. Cumulative production plots also find use in reporting for replanning purposes. (Refer ‘resource balancing’. bar charts and multiple activity charts contain essentially the same information. a bar chart becomes the same as a time-based cumulative production plot. namely via a network. where emphasis is placed on the different level controls that can lead to activity and project lengthening.13 (Continued).60 Conventional treatment from a systems perspective 12 11 10 9 8 Floors 7 6 4 3 2 1 0 Days (b) 5 art itio ns Int ern al p Figure 3. As Fin ish es Rou and gh-in s wet ervic area es s . Cumulative production plots. a resource schedule is one determined by resource numbers. The project decompression problem is a subproblem of the total planning problem. Decompression Decompression refers to the project duration being expanded. a network can be drawn showing the work involved. When ‘integrated’.

The causes of delays are many. Figure 3. so applying less resource numbers and decreasing the resource production rates increases (expands) an activity’s duration. A delay allowance. cumulative delay plot. In Figure 3. Its consensus meaning is that of end-product plus state final (terminal) conditions for the project. and so deliverables should not appear in a work breakdown structure. . Refer Carmichael (2004). However. Total delay allowance q Days Uniform rate of delay allowance usage 0 Project start Project completion Time Figure 3. it is best avoided. an end result has no meaning.14 is a project level representation. delay contingency or ‘time’ reserve may be incorporated into a project’s schedule where this is possible. Deliverable The term ‘deliverable’ is loosely used and abused. in spite of how useful a deliverable might be to some people for project management purposes. q days have been allowed over the total project. The diagonal line represents a uniform rate of usage of the delay allowance over the total project. In terms of work breakdown.14. or equivalent for a project milestone.14 Delay allowance. Delay A delay is a time period affecting the undertaking of or impeding the progress of a project. but only refer to those where uncertainty is involved. An expanded project duration may be possible if a later completion time (extension) is available or allowed. because of the term’s abuse.Planning tools and terminology A–Z 61 with the case of applying additional resource numbers or using higher resource production rates to decrease (compress) an activity’s duration.

In many projects. For an activity with a duration of d. End-product A project will commonly come about because of an identified need or want for some product. then d activity elements result. Earned value = (Estimated) Per cent complete × Planned cost Earned value is reporting at the project level. This distinction sometimes causes people confusion and many people are not aware of the distinction. The building is the end-product. facility. The processes that go together to materialise the building are the project. For example. . The earned value of production performed is found by multiplying the estimated per cent completion for the production by the planned cost for that production. This end-product is achieved through a project. earned value is the amount that should have been spent on the production done.). Refer Carmichael (2004). Earned value At any time during a project. Element An element of an activity represents a portion of an activity equal to the smallest chosen time unit for the project. people sometimes refer to a building as a project. service and so on. the smallest time unit is a day. It is not. However. Disturbance leads to the need for continual monitoring and reporting of project progress. but other time units are possible. or of the need to make a distinction. and possibly revised controls.62 Conventional treatment from a systems perspective Disturbance The corruption that prevents any planned for project output being obtained exactly is represented by disturbance. Error control Closed loop control may be referred to as error control if the control is selected based on the difference between desired system behaviour and actual system behaviour. asset. (It is acknowledged that the definition of a project is sufficiently flexible to include the operation and maintenance phase of a product within what is called the project. this is not the issue here.

Early on. These are used for tendering or replanning purposes. A common (but not the only) differentiation of degrees of accuracy in estimates is: • • • Approximate (within 20–25% accuracy). The cost of people. equipment and materials needed to carry out an activity is an example.Planning tools and terminology A–Z 63 Estimates One distinction between project costs is that between direct and indirect costs. If developing a tender price. the accuracy of estimates improves as the project stages progress. Indirect costs are one-off and time varying. or could be developed from first principles. Direct costs are costs that are directly proportional to the work quantity. reasonably accurate estimates are required. Detailed (within 5–10% accuracy). Price Margin/mark-up Estimated costs Direct costs Contingency Overheads Indirect costs Non-project-based costs Profit Project-based costs Figure 3. Examples include the salaries of project administrative staff. The usefulness of estimates depends on the skills of the estimator and the familiarity this person has with the work involved in the project.15 Tender price composition. Generally. estimates may be quite coarse. The total project cost is the sum of all the activity direct costs and indirect costs. hire of accommodation and insurance. These are compiled from overall unit rates for each piece of work in the project. and accrue whether or not any work is done. and are usually used for a decision whether to proceed with the project or not. . a contingency might be added to account for uncertainties. Several degrees of estimate accuracy may be recognised. (Also included would be head office overheads and profit [Figure 3. Preliminary (within 10–20% accuracy).15]. Usually used to assess the magnitude of the project and based on duration or cost indices. At implementation.) Estimates for durations and costs to carry out activities are usually based on past experience (extracted from historical records or ‘time’ studies) and trends. and they require either accurate unit rate methods or a bottom-up approach. The costs discussed in network analysis are generally direct costs.

as-constructed. an event has no meaning. availability of materials. in that it shows the actual durations to perform the various work items. as-implemented) programs. A factual network (as opposed to the original network) may not lend itself to analysis via conventional methods. such as milestone event. a similar situation can occur when start-to-start. A phase is started before the previous phase is complete. that is earlier delivery of the end-product. Fast-track project A project is said to be fast-tracked if its phases overlap (Figure 3.16). Event An event is the start or end of an activity. an occurrence at a specific time.64 Conventional treatment from a systems perspective Estimates depend on resource assumptions. but this can lead to misunderstanding. an event is not an activity. This can have the effect of completing the work sooner. in spite of how useful an event is for project management purposes. and earlier completion time for the project. The actual project cost could be expected to increase. Some computer packages may (for computational reasons) represent an event as an activity of zero duration. local and production factors including the availability of people and special skills. It is similar in concept to an as-built drawing. In terms of work breakdown. A factual network is a network (best done on a timescale) of as-executed activities indicating changes to the original program. Special or important events may be given names. including delays and variations. schedule extensions. It is a record of facts. Project phases are run in parallel where they may normally be run sequentially. liquidated damages and related claims. and agreed changes. Factual network Also called as-executed (as-built. And Nothing is as easy as it looks. and so events should not appear in a work breakdown structure. At the activity level. in place of more usual finish-to-start relationships. finish-to-finish or start-to-finish relationships are specified on the activity network. cross referenced to files on this information. in return for an earlier . which is the intent of fast-tracking. The factual network also finds a use in attempting to resolve disputes involving delays. According to Murphy’s Law (extended): Everything takes longer than you think. The factual network becomes a permanent record of a project’s progress. and availability of equipment/plant. The primary benefit of fast-tracking is a reduced schedule. Events correspond to the nodes in activity-on-link diagrams. weather and hazard facts.

is a more difficult task than budgeting for a business. The downside with fast-tracking is the associated peculiar managerial problems. compared with a business where next year’s operations could be expected to bear some similarity with this year’s operations. The term ‘budgeting’ may also be used to mean financial planning. Refer Carmichael (2004). Resource usage is converted to a unit of money. If this is related to an accounting system. return on money invested. Budgets tie expenditure to activities. . Projects are unique and have to be developed from first principles each time. for example earlier collection of rental.Planning tools and terminology A–Z 65 Phase 1 Phase 2 Phase 3 Phase 4 Conventional project phasing Phase 1 Phase 2 Phase 3 Phase 4 Fast-tracked phasing Figure 3. earlier sale and decreased cost of borrowed money. Budgeting for projects.16 Bar chart showing comparison of conventional and fast-tracked approaches. Though budgeting in a business via work packages (and later monitoring against these work packages) is little different to that involved in budgeting for projects. Financial planning Financial planning refers to all that involved in deciding the usage of money on a project. some might argue. and all other controls are held fixed. and the term ‘budget’ to mean financial plan. where emphasis is placed on the different level controls that can influence money usage (including recoupment). then as work is done. The financial planning problem is a subproblem of the total planning problem. this can be charged against individual account numbers.

a cumulative income plot can be superimposed on the cumulative expenditure plot to give Figure 3. Commonly project work is done by contractors. when invoiced. A positive cash flow is attractive to a contractor since it eliminates borrowing or tying up its own funds (which then become available for investment elsewhere). It may not be until the end of the project that the contractor shows a profit. when committed.17 Example cumulative expenditure and cumulative income diagrams (as-planned). income and cash flow. There may be large differences between cash flow patterns on different projects. including retention funds. For example. where a contractor (and similarly for a consultant) is being reimbursed by the project owner for work done. . or when paid. but not paid for by the owner until progress claims or similar are submitted. depending on the user. – have money being expended and various forms of income. subcontractor. The contractor funds the project. is received.17. Incomes and expenditures might be plotted based on dates of anticipated transfer of funds and billings. Negative cash flow draws on the contractor’s Cumulative money Profit Project expenditure (cumulative) Project income (cumulative) Progress payment Funds required to run project Time Project start Initial setting up of project Project completion Winding down of project Figure 3. Financial planning takes into account expenditure. Refer ‘S curve’. The vertical difference between the cumulative outlays of the contractor and the cumulative reimbursements represents the amount of money the contractor has to find on any day to finance the project. The comments below apply for both contractors and consultants.66 Conventional treatment from a systems perspective All the different participants in a project – owner. contractor. Many projects have a negative cash flow until the very end when final payment.

Receiving payment of a progress claim in a timely fashion. up-front payment. part payment on acceptance of tender. 60. and when it will be required to make progress payments (and hence by when money should be obtained). Progress payments may be invested and attract interest. subcontractors and suppliers to 30. Delaying payments to creditors. would generally be established based on a master plan level of generality. As shown in Figure 3. the cost of financing can be a major consideration. and reworked into the project. or similar terminology. Awareness of all these issues allows the contractor to put together a considered bid. The inflow of funds is also affected by profit gained on the project. items that occur early in a project are bid and reimbursed at marked-up prices/rates. there may be curiosity in the contractor’s cash flow. the contractor is funding the project except for right at the end of the project. Unbalancing its bid through front end loading. That is. Owners may try to defer payment to a contractor for as long as it does not hinder the contractor’s performance. The expenditure exceeds the income. or borrowing is necessary. From the owner’s viewpoint.Planning tools and terminology A–Z 67 working capital. For a project to proceed. or be selective in its bidding on projects. mobilisation payment. from an owner’s perspective. the owner may need to have an idea as to what its total costs for the project will be. The cost of financing On large projects. 90 or more days (a lag between when an invoice is received and when payment is made). but possibly more curiosity in the owner’s payment schedule which is the contractor’s income plot. implying a negative contractor cash flow. Feasibility. Planned expenditure will commonly lie somewhere between these two extremes. Starting activities closer to their LST rather than their EST. the owner is funding the project. Before awarding the work to a contractor. implying a positive cash flow for the contractor.17. advance payment. The ways that a contractor might improve its cash flow include: • • • • • Receiving a down payment. Contractors might attempt to improve their cash flows by moving the cumulative expenditure and cumulative income diagrams closer together. Interest on borrowed money can represent a significant contribution to the . The ideal from the contractor’s viewpoint is to have the cumulative income diagram above the cumulative expenditure diagram. funds availability has to exceed funds requirements. consultants.

. Cash flow calculations accordingly should include interest. but as free float in activity C. escalation and inventory costs. an escalation allowance is calculated for the midpoint of the time span over which the project work is expected to be performed. but this introduces inventory costs. It is called free float in C. expenditure tends to be postponed for as long as possible. (Commonly. critical activities have no float. Free float is a time period available. free float (FF) and interfering float (IF). Consideration of all these issues will lead to a minimum cost solution. Whether float is classified as free float or interfering float depends on where the activity occurs in the network. Float is a free time period. because if used in A and/or B. It is called interfering float in activities A and B. The top path has a combined duration of 3 + 2 + 4 = 9 days.68 Conventional treatment from a systems perspective project cost. the cheaper will be the associated end-product.) The earlier a project is completed. it affects the start time of C. The term ‘float’ is used in this book. but if used will affect (interfere with) the start times of following activities. it does not affect the start time of any other activity. Escalation is assigned to each work package relative to the duration of each package. giving a total float of 12 − 9 = 3 days. Several types of float may be distinguished. For example. and may be used in activities A. Float is the time period available to delay the start or finish of an activity without affecting the project duration. To guard against escalation. Therefore. The critical path runs through the bottom path and has a duration of 12 days. but if used will not affect the start times of following activities. TF = FF + IF The definitions of all floats are similar. because if used in C. Interfering float is a time period available. escalation costs are hard to predict. but only up to a combined total of 3 days. and end-product income can start to be received. Total float is the total time period available. subject to not having a deleterious effect on a project’s progress. Float Also called slack. It may be thought of as being composed of free float and/or interfering float. consider the part activity-on-node network of Figure 3. It is the same float manifesting itself in different ways. B and C in any combination. non-critical activities have float. damage and even theft while in storage. On the other hand. Conventionally. They are related through. which include deterioration. The most popular are total float (TF). components may be purchased early.18. This 3 days appears as interfering float in activities A and B. All are time periods available to delay the start or finish of an activity without affecting the project duration.

Gantt chart Refer ‘bar chart’. Where the float is negative. Lead time period refers to the duration between starting/finishing a preceding activity and starting/finishing a following activity. for example independent float and scheduled float. Activities are indicated by the letters A. B. and all other controls are held fixed. .Planning tools and terminology A–Z 69 TF = 3 FF = 0 IF = 3 A 3 TF = 3 FF = 0 IF = 3 B 2 TF = 3 FF = 3 IF = 0 C 4 D 12 TF = 0 FF = 0 IF = 0 Figure 3. where emphasis is placed on the different level controls that can influence the usage of people resources. Activity durations will have to be reduced by this float amount for the required completion date and calculated completion date to be the same. Refer ‘overlapping relationships’. and their durations are given within the circles. Human resource planning Human resource planning refers to all that involved in deciding the usage of people on a project. C and D. this implies that the required project completion date is sooner than the project completion date obtained through a critical path analysis. Lead time period Also called lead time and lag. Other floats are mentioned by various writers. The human resource planning problem is a subproblem of the total planning problem.18 Part-network illustrating float type. It is denoted LT in this book.

for example. Such information has direct relevance to planning any repetitive process. performance improves with a tapering off after carrying out the task many times. team or organisation gains experience in carrying out its allotted work. Typically the time period. the learning effect might be represented as in Figure 3. Figure 3. That distinction is not followed in this book. Learning As an individual. The curve is applicable for individuals. Knowledge of such possible improvements in performance can prompt a redesign of the work method. of course. then reducing the activity duration will reduce the associated lead time period.19 Learning curve. for example. does not happen where lead time periods are expressed as absolute values. to make or do a first unit is higher than the time period to make or do a second unit and this time period continues to decline for further units though to a slightly lesser extent. . in days. as a task is repeated.19 shows this schematically and is called a learning curve. Work/unit Units Figure 3. or as a percentage of the duration of an activity. When expressed as a percentage.20. this situation. increases. teams and organisations and has been substantiated in many field observations. Generally. Note that some writers make a distinction between ‘lead’ and ‘lag’. or converting processes into repetitive ones. the resultant time period savings may be incorporated into the plan or kept as a reserve. the effectiveness of the individual etc. In bar chart form. There is a saying ‘practice makes perfect’.70 Conventional treatment from a systems perspective Lead time period may be expressed as an absolute value.

The bars refer to the work involved with each unit.Planning tools and terminology A–Z 71 Unit 1 Unit 2 Non-learning Learning etc Unit 3 Figure 3.20 Bar chart representation of learning effect. this exponential curve becomes a straight line (Figure 3. Different values of b. The learning curve is commonly described by an equation of the form.21). represent different values of the rate of learning as shown in Figure 3. the rate of learning is specified as a percentage such as 70%. For example. . 80%. Conventionally.22. the index of learning.21 Learning curve on logarithmic scales. however. an 80% curve means that the work involved with the nth unit is 80% of that required for the T1 In (work/unit) In (Tn) ln(unit) ln(n) Figure 3. 90%. Tn = T1 nb where Tn average work/unit after n units T1 work/unit for the first unit b index of learning b < 0 When plotted on logarithmic coordinates.

1 b = – 0.22 Family of learning curves. Milestone A milestone is an important event occurring in a project. but nevertheless a milestone is an event not an activity.72 Conventional treatment from a systems perspective T1 Tn b = – 0. elements and constituents. to sub-subprojects.3 b = – 0. . 70%. Level A project may be decomposed to subprojects. or something should be accomplished by. . It is a process-oriented approach which has its origins in the manufacturing industry. A project may have a number of milestones. to activities. some computer packages may define a milestone by setting a bogus activity’s duration to zero. Line of balance technique The line of balance technique applies to resourcing on projects that typically have sequential/repetitive work content (for example. Milestones are shown on a network or bar chart in order that everyone is aware when particular important events occur. Typically. n − 1 th unit. As a computational and representational device.5 n Figure 3. the work of trades in constructing a high-rise building). Both T1 and b require estimating. 80% and 90% curves correspond to b values of about −0 51 −0 32 and −0 19 respectively. Refer ‘resource balancing’. and ‘cumulative production plot’. Levels refer to the strata in such a hierarchical decomposition. these are adapted from similar situations or historical records.

while the fourth step is bookkeeping. is a numerical probabilistic method of network analysis. Multiple activity chart Also called a multi-activity chart. The currently popular way to realistically incorporate variability in activity durations (typically represented by probability distributions) for network analysis is to use Monte Carlo simulation. Separate bars are drawn for each resource. Some commercially available Monte Carlo simulation packages sit as overlays on critical path method (CPM) packages. it is a chart on which the activities of more than one resource (people or equipment) are recorded on a common timescale to show their interrelationship. Monte Carlo simulation is suited ideally to computer use. the second step is analysis. . A conventional network analysis (incorporating overlapping relationships if applicable) is carried out. note that Monte Carlo simulation is a technique for analysing nearly anything that contains probabilities. However. as used in planning. are collected. The above two stages are repeated many times. The first step is one of data generation. such as project completion times.Planning tools and terminology A–Z 73 Monitoring Monitoring involves measuring or observing the output or performance of a project. Monte Carlo simulation essentially converts a difficult probabilistic problem into many simpler deterministic problems. Relevant statistics on important items. It is not a technique for use by hand or where a closed-form solution is required. The steps involved for network analysis are: • • • • Activity durations are sampled (via the generation of uniform random numbers). A multiple activity chart is drawn in a similar sense to bar charts with a timescale and shading length in proportion to the duration of whatever is being plotted. where large amounts of data can be handled with ease. The bars may be drawn vertically (top to bottom time sense) or horizontally (left to right time sense). Monte Carlo simulation Monte Carlo simulation.

bar charts and multiple activity charts contain essentially the same information. Times for such charts are typically based on field records.23 Example multiple activity chart. apart from that. Bar charts tend to be developed beforehand.74 Conventional treatment from a systems perspective Machine Operator Supervisor Time Not involved in work Involved in work Figure 3. but there is no reason why connections cannot be included. . Cumulative production plots. The chart shows very readily periods of idleness and periods when the resource types are utilised. Murphy’s Law The origin of Murphy’s Law.23 shows an example of multiple activity chart. they are no different. Figure 3. usually without connections. while multiple activity charts tend to show what has happened. Some of the published attribution is not convincing. Possibly the naming came about because Murphy is a distinctly Irish name. the original principle ‘If anything can go wrong it will ’ is unclear. A multiple activity chart is essentially a different form of bar chart concentrating on what particular resource types are doing.

Methuen. There are many extensions and corollaries of the basic principle behind Murphy’s Law. O’Toole’s commentary on Murphy’s Law Murphy was an optimist. Nagler’s comment on the origin of Murphy’s Law Murphy’s Law was not propounded by Murphy. Nevertheless. but by another man of the same name. Many of the following examples are from Murphy’s Law. some tasks are completed. Sodd’s First Law When a person attempts a task. he or she will be thwarted in that task by the unconscious intervention of some other presence (animate or inanimate). ends badly. since the intervening presence is itself attempting a task and is. Goldberg’s commentary O’Toole was an optimist. Murphy has something to say on planning related matters. you have to do something else. the greater the alterations required. Nothing is as easy as it looks. Before you can do something. All give a humorous view of laws guiding life. ends worse. London.Planning tools and terminology A–Z 75 and the Irish are known for their quirky humour. 1986. Some examples are: No matter what goes wrong. and again the authorship of these is unclear. Anything that begins badly. A. . it will probably look right. The reliability principle The difference between the Laws of Nature and Murphy’s Law is that with the Laws of Nature you can count on things screwing up the same way every time. Everything takes longer than you think. Now they tell us Law Information calling for a change in plans will arrive after the plans are complete. of course. Pudder’s Law Anything that begins well. Bloch. The nearer to completing the job. subject to interference.

Corollary. and change the unit of measure to the next highest unit. unexpected delays. 4. Patton’s Law A good plan today is better than a perfect plan tomorrow. Frothingham’s Fallacy Time is money. Workshop Principle (number 4) The more carefully you plan a project. 2. Cheop’s Law Nothing ever gets built on schedule or within budget. if two projects must be done at the same time they will require the use of the same part of the work space. . The Einstein Extension of Parkinson’s Law A work project expands to fill the space available. The carelessly planned project takes three times longer to complete than expected. the more confusion there is when something goes wrong. No matter how large the work space.76 Conventional treatment from a systems perspective Wynne’s Law Negative slack tends to increase. Fuzzy project [goals] are used to avoid the embarrassment of estimating the corresponding costs. 3rd Corollary to Law of Applied Confusion After adding two weeks to the schedule for unexpected delays. Project teams detest weekly progress reporting because it so vividly manifests their lack of progress. a carefully planned project takes only twice as long. 3. Westheimer’s Rule To estimate the time it takes to do a task: estimate the time you think it should take. Thus we allocate 2 days for a one-hour task. The effort required to correct course increases geometrically with time. multiply by 2. and the last ten percent takes the other ninety percent. Ninety–Ninety Rule of Project Schedules The first ninety percent of the task takes ninety percent of the time. Laws of Computerdom according to Golub 1. add two more for the unexpected.

Planning tools and terminology A–Z 77 McDonald’s Corollary to Murphy’s Law In any given set of circumstances. With this calendar an order given on the 7th can be carried out on the 3rd. . There are no bothersome non-productive Saturdays and Sundays. Kent Family Law Never change your plans because of the weather. So there are two Mondays in each week. Everyone will be happy and ulcer free. Seay’s Law Nothing ever comes out as planned. the proper course of action is determined by subsequent events. ‘while you wait’ and ‘same day’ jobs may be handled without interruption to other promises. There are several extra days at the end of the month for those end of the month rushes. Acceptance of this calendar is made easier by pointing out that this system has been in unofficial use in many companies for some years. On this day. Sweeney’s Law The length of a progress report is inversely proportional to the amount of progress. Everyone wants things done yesterday. FRI 7 14 21 28 35 MON 6 13 20 27 34 TUES 5 12 19 26 33 WED 4 11 18 25 32 THURS 3 10 17 24 31 MON 2 9 16 23 30 SAME DAY 1 8 15 22 29 This calendar has certain advantages: • • • • • • Every job is a rush job. First Law of Corporate Planning Anything that can be changed will be changed until there is no time left to change anything. There is a new day each week – ‘same day’. The following Murphy’s Calendar is useful for planners. Everybody wants things done early – on Mondays for preference.

or the logic of how the project is to be undertaken. represent a rational intermediary step to project programs.25a). In drawing networks. • Terminology used in this book is activity-on-link diagram and activity-onnode diagram. with arrows linking the activities (Figures 3.78 Conventional treatment from a systems perspective Network A network is a diagram showing the logical sequence of undertaking activities. In general. In other words. . (a) Activity-on-link diagram. For networks used in planning.24 Network diagram types.25b). resource and expenditure plots and S curves. even though this is not the practice in other publications which generally use conflicting terminology. the network may accordingly be called a directed network. Project activity (a) Project activity (b) Figure 3.24a and 3. the work order interrelationship of the activities.24b and 3. it shows the precedence relationships between activities. and both are collectively referred to as precedence diagrams. the links are directional and are represented by arrows. This gives rise to the two fundamental types of networks: • Activity-on-link diagrams (elsewhere called arrow diagrams or activityon-arrow diagrams) whereby activities are represented by arrows starting and finishing at nodes (also called starting and finishing events) (Figures 3. a network is composed of links and nodes. Activity-on-node diagrams (elsewhere called precedence diagrams or circle diagrams) whereby activities are represented by the network nodes. care has to be exercised that false logic is not introduced. as part of an iterative analysis approach to planning. Networks are the basis of project planning and. (b) Activityon-node diagram.

Planning tools and terminology A–Z 79 Approval. Figure 3. and time-scaled network. Figure 3. that is.1. dsgn base cpts Manuf. options Assemble options Apply finishings Det. Activity-on-link diagrams were the basis of early critical path analysis work in the 1960s and 1970s. (a) Activity-on-link diagram. Some computer packages allow both types of diagrams. It is possible to develop a transformation between the two types of diagrams. design Approval.) The same analysis is used for both network types. options Det. . In favour of activity-on-link diagrams. and the two types of diagrams give the same answers. base cpts Prelim.55. Recent favour by users seems to be towards activity-on-node diagrams.25 Example networks. dsgn. Clients may also insist on the use of a particular computer package which in turn will often determine the type of network used. design Approval. although the orientation of computer packages accessible to the user is an influence. Many unknowingly use activity-on-link diagrams (to a timescale) when using certain software packages. options Apply finishings Marketing Sales (a) Approval. See the associated bar chart. and activity-on-link diagrams are the only type that can be represented to a timescale. both activity-on-node and activityon-link diagrams are being used simultaneously. Some people insist that activity-on-node diagrams are much better than activity-on-link diagrams. base cpts Deliver base cpts Assemble base cpts Marketing Sales (b) Figure 3. dsgn. base cpts Dummy Deliver base cpts Assemble base cpts Assemble options Prelim. bar charts are universally used. (b) Activity-on-node diagram. options Dummy Det. base cpts Det. It is a matter of personal preference which network type it used. the activity-on-link diagram might be disguised as a connected bar chart – this is in fact a time-scaled activity-on-link diagram. dsgn base cpts Manuf. (Activity names have been abbreviated because of space limitations – work items are implied. yet the same people use connected bar charts to convey a project’s program.

A Storey 2 C Dummies A Storey 3 B C Dummies etc (b) Figure 3.26 Part activity-on-node and activity-on-link diagrams applying to the construction of a multistorey building. (Activity names have been abbreviated because of space limitations – work items are implied. For overlapping relationships. Adding and taking away links in activity-on-node diagrams tends to be easier than adding and taking away nodes in an activity-on-link diagram. An activity can be connected to other activities directly. (b) Activityon-link diagram. they allow easier modification (editing or updating) once initially developed. The activity-on-node diagram tends to be easier to understand and follow by all strata of management and permits an easier numbering or coding scheme. (a) Activity-on-node diagram. Storey 2 Storey 3 etc (a) Trade A Storey 1 Trade B Trade C Dummies B etc. the vertical arrows between storeys are dummy activities.80 Conventional treatment from a systems perspective In favour of activity-on-node diagrams.) . with an activity-on-link diagram there may be a need to introduce a dummy activity in order to not affect the required logic. Trade A Storey 1 Trade B Trade C etc. such relationships apply between activities and these are easier to represent on an activity-on-node diagram.

In drawing a network: • • • The diagram is not drawn to any scale. which relates to the construction of a multistorey building. Where numbering (optional) is used for links and/or nodes.26. Redundant logic can be removed. A following activity may depend on a preceding activity(ies) according to any of four overlapping relationships – finish-to-start (F/S). railway lines and like structures involving sequential/repetitive activities. and that is where there are the same sequential activities repeating. activity-on-node diagrams may also be preferred. and also to locate the important activities near the (vertical) centre of the page. but are not essential. each trade in sequence works its way up the building. See for example. The size and form of the shapes are not important. although a computer package may require a specific convention. Refer Carmichael (1989). Figure 3. The activity-on-node diagram turns out to be a more economical representation compared to the activity-on-link diagram. the numbering scheme can usually be anything the user chooses. Multiple dependency gives rise to multiple links into and/or out of an activity (activity-on-node diagram) or multiple activities into or out of nodes (activity-on-link diagram). It may be found convenient though to locate activities associated with a certain trade or function together. for example relating to roads. Arrows do not have to be straight. start-tostart (S/S). Various drafting styles are possible for nodes. • • • • • • • • • . pipelines.9 shows some examples. pipeline style drafting can be used to help reduce possible confusion. The diagram has a left to right sense. Activities to the left of the diagram could be expected to occur before activities to the right of the diagram. which needs the incorporation of many dummy activities to maintain the logic. Activities that are dependent are drawn ‘in sequence’. Multiple relationships may also occur between activities. For other ‘linear’ type projects. Dummy ‘Project Start’ and ‘Project Finish’ activities can be added for completeness. Where arrows cross over. The length and direction of the arrows is not important. Figure 3. Activities that can be done concurrently are drawn ‘in parallel’.Planning tools and terminology A–Z 81 There is a special case where activity-on-node diagrams are preferred to activity-on-link diagrams. and project end is at the right. Circular logic or circuits are not possible. ranging from circles to multicelled rectangles. but generally have a left to right sense. Activities can be located anywhere vertically on the page. finish-to-finish (F/F) or start-to-finish (S/F). where the project start is at the left.

Commonly. The criteria by which the preferred end-product is selected are the end-product objectives.27 End-product and project objectives. The materialisation of a project’s end-product can be performed. ‘PERT’ and ‘Monte Carlo simulation’. Form. Work method. Objectives The plural terms ‘objectives’ and ‘constraints’ are generally used in this book even though the singular forms may be applicable in some situations. project duration and deviation from specification. End-product objectives 2. CPM’ (equivalently critical path analysis. minimum Objectives End-product objectives Project objectives Hierarchical levels of objectives Figure 3. In systems studies terms. resource and resource production rate considerations on projects follow. the problem is an inverse problem. of the endproduct follow. scope. Project objectives (Figure 3. on any project there are two types of objectives: 1. The selection of the preferred end-product form is a design problem. there are possibly an infinite number of versions of end-products. and hence there is no unique solution. but other objectives are possible. possibly. the exception is where the singular is deliberately intended. function. In a similar idea. And these may apply throughout the project (for example. project objectives say something about project cost. . The criteria by which the preferred materialisation of the end-product is selected are the project objectives. CPA). finishes etc. That is.82 Conventional treatment from a systems perspective Network analysis Refer ‘critical path method.27). The selection of the preferred materialisation or means to the end-product is the solution to the planning problem. in an infinite number of ways.

cyclones. ) Social impacts Geotechnical considerations. general project objectives will contain a component over the time domain of the project and a component at the final (terminal) point. . minimum duration which is related to the project completion time). These are expressed in terms of end-product matters or project matters. for example originating from a corporate plan. element and constituent levels. . ) ˙ Resource usage Quality issues Community acceptance Environmental (natural) effects Safety Risks Public impact Extreme event impact (floods. concerns. As needed to perform systematic problem solving/systems synthesis. or at the final (terminal) point of the project (for example. activity. environmental (natural). to: • • • • • • • • • • • • Money (end-product – sales.Planning tools and terminology A–Z 83 deviation from specification). Objectives may be expressed at the various project. benefit:cost ratio (BCR). net present value (NPV). for example. project – cost. project – duration. alternative names include: • • • • • • • (Optimality) criterion Performance index Performance measure Merit function Payoff function Figure of merit Aim . marketing. as the case may be. Refer Carmichael (2004). End-product objectives and project objectives (and constraints) may relate. budget. Alternative names for objective. Such values may also reflect political. ) Duration (end-product – lifetime. That is. End-product and project objectives may derive from higher level values within an organisation.

and finish-to-start (F/S) relationships with non-zero lead time periods (LT) (Figure 3. while a finish-to-finish relationship implies the following activity cannot finish until some specified time period after the previous activity has finished. The activities are said to have finish-to-start (F/S) relationships. namely start-to-start (S/S). and no follow-up monitoring of system performance and corrective control or action is carried out. For the stripped-back projects considered in this book. Overlapping relationships An elementary development of networks assumes that when one activity finishes another can start (with zero lead time period between the activities). These generally occur outside the management literature. S/S. They are common in the optimisation. Planners seem to have most usage for F/S. For example. finish-tofinish (F/F). output and state are the same. activities run in parallel instead of sequentially. performance or behaviour. Output contains multiple parts (that is. F/F and S/F activity overlapping relationships can be like fast-tracking at the activity level. Using S/S. 1981). it is a vector quantity of output variables). a start-to-start relationship implies the following activity cannot start until some specified time period after the previous activity has started. optimal design and optimal control literature (Carmichael. Output The output describes the external or observable system response. that is there are no observability (in the sense of Kalman) issues.28). Open loop control In open loop control. start-to-finish (S/F).84 Conventional treatment from a systems perspective • • • • Goal Cost function Design index Target function. the control is selected up front. F/F and S/F relationships in that order. Output is a controlled variable. Generalisations of this are possible to embrace various relationships between the starts and finishes of dependent activities. .

for example. overlapping relationships can be avoided through the use of subactivities. Figures 3. See for example.29–3. which relates to the construction of a . Example A situation in which start-to-start relationships.31. bar chart representation. and also economical in terms of representations such as the network diagram. However.Planning tools and terminology A–Z 85 LT Finish-to-start (F/S) with lead time period LTi Start-to-start (S/S) with lead time period LTj Finish-to-finish (F/F) with lead time period LTi LTj Start-to-finish (S/F) with lead time period Figure 3. find application is illustrated in Figure 3.32. Subactivities Many people find the overlapping facility helpful. bar chart and so on.28 Possible overlapping relationships between activities.

(d) Alternative drafting style. For other ‘linear’ type projects. (a) Bar chart.86 Conventional treatment from a systems perspective Time i Starting trade (Str) Following trade (Ftr) j LTi = 3 days (a) First 3 days Part Str Remainder Part Str Ftr (b) LTi = 3 days Str (c) Ftr Str LTi = 3 days (d) Ftr Figure 3. relating to roads. Start-to-start relationships may apply between trades in order to accelerate the completion of the building. (c) Using overlapping notation. (Activity names have been abbreviated because of space limitations – work items are implied. railway lines and like structures. pipelines.29 Start-to-start relationship example. start-to-start relationships may also be preferred. . (b) Without using overlapping ideas. each trade works its way up the building. it may not be necessary for one trade to completely finish on a floor/storey before the following trade commences.) multistorey building. for example.

For example. (d) Alternative drafting style. (a) Bar chart. Figure 3. (b) Without overlapping ideas. for example.Planning tools and terminology A–Z 87 Time i Preparation (Prep) Finishing treatment (Ftr) j LTj = 2 days (a) Second part Prep Part Ftr Part Ftr First part (b) LTj = 2 days Prep Ftr (c) LTj = 2 days Prep Ftr (d) Figure 3.33b introduces a new S/F relationship between the first and last activity.30 Finish-to-finish relationship example.33. which relates to imposing a completion time limit dependent on the start time of the first activity.33a shows the original network.) Example A situation in which start-to-finish relationships. (Activity names have been abbreviated because of space limitations – work items are implied. the start of the first activity . Figure 3. find application is illustrated in Figure 3. (c) Using overlapping notation.

Example In constructing a building. (d) Alternative drafting style.88 Conventional treatment from a systems perspective Time LTi = 3 days i Remove trees etc Construction and replant trees j LTj = 2 days (a) First 3 days Part remove Remainder Part remove Const. (Activity names have been abbreviated because of space limitations – work items are implied. (b) Without overlapping ideas. (c) Using overlapping notation. each floor follows the next. Figure 3./ replant (c) LTi + LTj = 3 + 2 = 5 days Remove trees (d) Const. .34 shows part of the construction in bar chart form.) and the end of the last activity may represent milestones constrained by a given time interval between when they occur.31 Start-to-finish relationship example. For example. (a) Bar chart. Replant trees (b) LTi = 3 days Remove trees LTj = 2 days Const./ replant Figure 3.

3 F/S Act. 3 F/S Act.32 Part activity-on-node diagram applying to the construction of a multistorey building. conventional finish-to-start (F/S) relationships are used.) Act. The S/S relationship dominates between the second and third activities. 4 (a) Act. The second activity has float at its start because the F/F relationship with the first activity dominates. 3 Act. 4 Act. Figure 3.33 Example usage of start-to-finish relationship. 2 F/S Act.34a.34b is sometimes termed a ladder representation. (Work items implied. 1 F/S F/S F/S S/F (b) Act. 2 Act. 1 F/S Act. 1 S/F F/S Act. In Figure 3. .Planning tools and terminology A–Z 89 Trade A Storey 1 F/S Storey 2 S/S Trade B S/S F/S Trade C S/S F/S Trade D S/S F/S S/S F/S Storey 3 S/S F/S Storey 4 F/S F/S F/S S/S F/S S/S F/S S/S S/S F/S S/S F/S S/S S/S F/S S/S F/S S/S F/S S/S Figure 3. 2 F/S Act. 4 Figure 3.

(b) Using S/S and F/F relationships. 80% of a project cost estimate is the result of 20% of the component cost items. 80% of a person’s output is due to 20% of that person’s tasks. Parkinson’s Law Parkinson’s (first) law Work expands so as to fill the time available for its completion (Parkinson’s Law or the Pursuit of Progress. building construction.) Pareto rule/principle The Pareto rule/principle is named after Vilfredo Pareto (19th century). The numbers 80 and 20 are not to be interpreted precisely. Originally applied to wealth distribution. Northcote Parkinson.90 Conventional treatment from a systems perspective 40 days Ground floor slab 12 days Columns 40 days First floor slab 0 (a) 40 52 Days 92 Ground floor slab 40 days F/F lead time period 3 days Columns 12 days F/F lead time period 10 days S/S lead time period 10 days S/S lead time First floor slab period 3 days 40 days 0 (b) 31 34 40 43 Days 74 Figure 3. The thing to be done swells in perceived importance . ( Work items implied. It is sometimes called the 80:20 rule/principle. it has since been generalised: 80% of the outcomes (outputs) are the result of 20% of the influences (inputs). (a) Conventional F/S relationships. There are many applications in project management. C. London.34 Example. 1957). John Murray. For example.

A performance measure. PERT PERT (acronym for program evaluation and review technique) is a probabilistic first-order method of network analysis where the duration of each activity is described by a probability distribution derived from an optimistic estimate. A performance measure is in the lay usage sense (but not in an optimal systems sense) of the word ‘objective’ (mentioned in Carmichael. from any of its sources. objectives are developed very early. 2004). but all of these types of terms are used very loosely by most people. The associated calculations enable the probabilities of completing events by certain times or keeping to schedule to be calculated. could be used as the basis for error control thinking. A performance measure may also be based on some industry.Planning tools and terminology A–Z 91 and complexity in a direct ratio with the time period to be spent in its completion. Performance measures tend to be developed later in projects (after the initial planning has been done). A performance measure could be expected to follow planning baselines or targets (as-planned values) for resource usage (and cost and production).35). The analysis by PERT is very similar to that involved in the critical path method (CPM – a deterministic method where the durations of activities are estimated to single values) calculations except that PERT carries along one . Parkinson’s law of delay Delay is the deadliest form of denial. The terms are used frequently by project personnel because they sound good and impress. A performance measure might be termed a key performance indicator (KPI) or similar. Performance measure Performance measures are gauges by which the success or otherwise of a project is measured. This information may be used to schedule resources so as to reduce the possibility of delays within a project. a pessimistic estimate and a most likely estimate (Figure 3. competitor or previous project benchmark. They allow the ranking of activities (an indication of their criticality) according to the probability that the free time period at events is greater than or equal to zero. Parkinson’s (second) law Expenditures rise to meet income. but lack precision.

that can give the misleading impression that PERT uses an activity-on-node diagram as its basis.35 Probability density function for activity duration. PERT uses an activity-on-link network although events may be given descriptions. PERT appears in nearly every textbook on project planning because the technique is useful as a teaching tool to make planners more aware of the assumptions on which deterministic network calculations are based. of distinguishing events along with concentrating the calculations on event times. and compressing networks in a probabilistic sense raises its own issues. PERT gives not only an expected project completion time but also the scatter that may be anticipated in the project completion time. PERT calculations work with an activity expected duration and a measure of the scatter or variability in the duration (the variance). Probabilistic network analyses appear to be best handled using the technique of Monte Carlo simulation. Planning is an inherently probabilistic problem although in practice it may not be treated as such. namely the activity duration variance. the project requires compressing. Many people and some computer packages use the term ‘PERT’ but incorrectly – they are in fact meaning CPM. CPM might be considered a special (deterministic) case of PERT with a couple of differences. additional quantity in the calculations. There is much abuse of the term ‘PERT’. It is this characteristic. (Variance is used here in the probabilistic sense. However. PERT is only developed on activity-on-link networks.) No one appears to be using PERT in practice because such probabilistic approaches are not popular. a single project completion time is obtained. It is thus possible that earlier completion times could be obtained (without the need for compression) if activity durations occur closer to their optimistic durations rather than . That is. When the deterministic calculations of the critical path method are carried out. To achieve lesser completion times.92 Conventional treatment from a systems perspective Frequency of occurrence Most likely duration Optimistic duration Pessimistic duration Activity duration Figure 3.

in what order and when and with what resources (type. When network compression is required. Precedence Precedence refers to the logical connection or dependence between activities as indicated by a chosen work method. it may be carried out similarly to the deterministic case.36 indicates the same situation for a single activity but the concept is applicable for both activity durations and project durations.36 Crashing an activity showing mean times mt . planning Planning establishes how and what work will be carried out. Although earlier project completion times may have a finite probability.Planning tools and terminology A–Z 93 Frequency of occurrence Crashed activity (c) Compressed activity (p) Normal activity (n) mc t mp t mn t Time Figure 3. additionally expressed in a money unit). resources and resource production rates) throughout the project duration. their most likely durations. some compression must be carried out. A plan (noun) is the outcome of planning. A similar argument applies to completion times greater than the expected project completion time. . network compression using PERT is best used as a guide as to where to apply attention in the network rather than giving absolute results. and number or quantity. To plan (verb) is the act of choosing the controls (method. To ensure a higher probability of achieving an earlier project completion time. Plan. The situation in Figure 3. The dependence may be to the start or finish of a whole or part activity. Corresponding to each expected project completion time there is a variance. In general. Cost slopes may be defined in terms of activity expected durations and associated costs. this probability gets smaller as the project completion time gets earlier.

Process charts Process charts portray a sequence of activities diagrammatically by means of a set of process chart symbols to help a person visualise a process as a means to examining and improving it (Currie. The resulting network diagram is no more than the answers to Q1. Which activities can be done at the same time as this activity? Answers to these questions establish the precedence relationships between activities. in whole or in part: Q1. Generally. A flow process chart sets out the sequence of the flow of a product or a procedure by recording all activities under review using appropriate process chart symbols. and the logic of how the project is to be put together.37. keeps or retains Interferes or delays Verifies quantity and/or quality ∇ D STORAGE DELAY INSPECTION Figure 3. Further breaking down of the work may be possible and is sometimes done. then the network diagram will have activity B connected from activity A. 1969). 1969). An outline process chart gives an overall picture by recording in sequence only the main operations and inspections.94 Conventional treatment from a systems perspective Precedence may be established. The work is broken down into its component activities which are represented schematically by these symbols. Q2 and Q3. and represents the logic of how the project is to be carried out. furthers the process Travels Holds. 1959. ILO.37 Symbols for process chart construction (Currie. if one (B) activity’s progress is dependent on another (A) activity’s progress. Refer ‘network’ and ‘overlapping relationships’. 1959. A resource type flow process chart records how the resource is used. accomplishes. Symbol Activity OPERATION TRANSPORT Predominant result Produces. . Process charts are constructed using standard symbols defined in Figure 3. by addressing the following questions for each activity. amongst other ways. Which activities does this activity follow or logically depend upon? Q2. Which activities does this activity logically precede? Q3. ILO.

DELAY and STORAGE activities. References Currie.39 shows an example flow process chart. loops and multiple parallel paths respectively. A method study is carried out before any work begins. only the OPERATION and INSPECTION symbols are used. when work is divided or reprocessed. The chart may be drawn from the point of view of the worker or the equipment or material used by the worker. Activities can be numbered in order. M.Planning tools and terminology A–Z 95 An outline process chart is useful in design and planning. Sir Isaac Pitman and Sons Ltd. Activity durations can also be placed on the chart. such variants lead to branching. . Figure 3. are introduced. Figure 3. for example.38 Example outline process chart. Variants on the theme in Figure 3. Work Study.38 occur. or alternative means are possible to perform an activity. It does not show who carries out the work or where the work is carried out. Geneva. and descriptions are placed adjacent to the symbols. Main work Peripheral/ subsidiary work Description " " Description " " " " " " Figure 3. London. An Introduction to Work Study. Figures 3. information etc. ILO (International Labour Office) (1969). (1959). The chart gives an overall view of the work and graphically shows the sequence of activities where inspections occur and where extraneous materials.40 shows an alternative flow process chart. Travel distances can be included to scale or by annotation. International Labour Office. R. A flow process chart extends the outline process chart to include TRANSPORT. It is similar in appearance to the outline process chart. Accordingly.38 shows an example outline process chart.

Activity Wait Load Check Manoeuvre Travel Wait Unload Check Stockpile ∇ Remarks Total 2 2 1 2 2 Figure 3. production rate ‘Production’ is used in this book in a general sense to indicate some measure of work done. m or items. and the sequence of activities needed to achieve the desired project end point. A program contains information at both the activity and . Production rate is this work done per unit of time. It gives the time relationship of activities. for example m2 . Program A program or schedule conveys when work is to be done (and hence it follows. or when combined with resource numbers to that achieved in an activity or project. Production. when resources and money are needed). The unit of measurement of production can be various.96 Conventional treatment from a systems perspective Wait Load Manoeuvre Check Check ∇ Stockpile Unload Wait Travel Figure 3. Production rate may refer to a resource’s constituent behaviour.39 Example flow process chart to emphasise distances travelled.40 Example flow process chart.

with starting and ending points. and the replanning loop is not closed. A project is a system and a system can be decomposed to interacting subsystems which are themselves systems. which is itself a project. throughout a project. Amongst planners. A program is presented for the purpose intended. . Subprojects. Refer Carmichael (2004). and with defined objectives and constraints. The use is also not to be confused with a ‘program’ referring to a collection of projects. They typically derive from a network analysis. 2004). Each subproject. enlargements upon this meaning will be found in the literature. including that they are: • • Unique (one-off. The user has a need for the quick retrieval of relevant information. though perhaps unsatisfactory in a number of ways.41). the term ‘program’ or ‘schedule’ may be applied to any or all of the following: • • • • • (Connected) bar chart Time-scaled network diagram Cumulative production plot Activity date/time listings/timetable Marked-up drawings. specific discrete undertaking with a unique environment and unique constraints). Sometimes bad practice is seen in that the program becomes an end in itself. Perhaps a more suitable way of thinking about a project would be in terms of attributes that are characteristic of projects. Finding a satisfying definition of a project is difficult. looks after a part of the total project.Planning tools and terminology A–Z 97 project levels. The definition. or set of activities. Project A good definition of a project is hard to find. and resource consumption’ is useful. Programs show referenced (as-planned) baselines against which actual progress can be compared. Programs are continually updated. However. ‘Any undertaking. (Carmichael. computer programs (commercial software packages) may however be used to generate project programs. Likewise a project can be decomposed to interacting subprojects (Figure 3. that is a project is a subsystem of a program. The use of the term ‘program’ here is not to be confused with a computer program. Finite (definable start and end points). to reflect the latest information available.

(ii) A management function approach. Many published definitions of project management are partly circular describing management as management.98 Conventional treatment from a systems perspective Project Subproject Sub-subproject Figure 3. Management is commonly described in general management texts in terms of: • • • • • • ‘Planning’ Organising Staffing Directing ‘Controlling’ Coordinating. Most definitions are usually in terms of what a project manager does. 2004).41 Decomposition of a project. They often come from studies made of actual projects and observations carried out on the project managers. There are three favoured ways of describing what a project manager does. Project management A good definition of project management is hard to find (Carmichael. but other ways do exist: (i) A ‘classical’ management approach. An alternative view of project management is to regard it as a collection or integration of subfunctions: • • • Scope management Quality management ‘Time’ management .

many people feel comfortable with a chronological way of describing a project and the activities involved. Replanning Replanning gives revised values of the control variables (method. ‘Controlling’ in (i) above is in the loose lay meaning sense of the word. The form of reporting is chosen to match the intended recipient. Reserve Also called contingency. Reporting Reporting refers to communicating the results of project monitoring. The term ‘replanning’ is preferred to ‘project control’ in this book. A project may be seen as going through a number of phases. There is no consensus as to the naming of phases or to the number of phases. project output or performance is reported. (iii) Chronological approach.43). Reporting can take many forms. perhaps in a summarised form. resources and resource production rates) for the remainder of the project. That is. materials include parts and components. Others find the first two breakdowns so unsystematic as to be of no use. They are also distinguished from materials which refer to that which is incorporated into whatever is being produced. Some people find such breakdowns useful. . In terms of management. because lay sloppy usage of the latter has a tendency to mislead.Planning tools and terminology A–Z 99 • • • • • Cost management Risk management Contract/procurement management Human resources management Communication management. A reserve is an extra amount (in money or time period) added to an initial cost or duration estimate respectively to take care of possible unknowns. Two types of reserve are commonly used – cost reserve and ‘time’ reserve.42 and 3. They are distinguished from money which is a common unit of measurement of resource usage (Figures 3. Resource Resources refer to people and equipment/plant undertaking the production. and is inconsistent with the optimum control systems usage preferred in this book.

with or without a buffer between them (Figure 3.42 Preferred reference to resources. the terminology ‘resources’ may be used to refer to resources generally. Some people refer (inappropriately in planning) to time. Project duration is considered a higher priority than project cost. Two extremes in handling resources are possible: 1. and typically resourcing sequential/repetitive work.43 Some possible relationships between resource usage and money. Resource numbers (quantity) are chosen to give a desired activity production rate. skills etc.100 Conventional treatment from a systems perspective Resources People Equipment Common unit of measurement Money Figure 3. it may be desirable to have the same activity production rates for each activity. Where multiple activities occur (sequentially or otherwise). plots for each activity on a cumulative production plot are parallel. as resources. In this book. to resource numbers (quantity) or resource types. Money Resource usage Figure 3. under such circumstances. Resource balancing ‘Resource balancing’ is a term applied to resourcing work. overheads. The term ‘resource types’ is used where a distinction between the various resources is considered necessary.44). . The term ‘resource number (quantity)’ is used where a distinction within a resource type is considered necessary.

and the project completion date will relate to the lowest activity production rate (Figure 3. and S/S relationship.Planning tools and terminology A–Z 101 2. Where multiple activities occur (sequentially or otherwise).45 Example activities advancing at different rates. .45). the start or finish times of activities are adjusted in order to avoid interference between activities. and F/F relationship. plots for each activity on a cumulative production plot are not parallel.44 Example activities advancing at the same rate. 20 days S/S lead time period 3 days 20 days Excavate trench Lay pipe Excavate trench Chainage Lay pipe 0 3 Days 20 23 Figure 3. The production rate for each activity is determined by the available or usual resourcing numbers (quantity). 20 days Excavate trench Lay pipe 12 days F/F lead time period 0 days Excavate trench Chainage Lay pipe 0 8 Days 20 Figure 3. Project cost is considered a higher priority than project duration.

Labourers Days Carpenters Days Figure 3. . A resource plot gives resource requirements versus time.102 Conventional treatment from a systems perspective Constituent level and element level controls are adjusted to give desired activity and project level performance. Resource plot Also called a resource profile or resource histogram. The resource planning problem is a subproblem of the total planning problem. over the duration of the project. The second method might be termed ‘resource scheduling’. this thinking may be called the line of balance technique. An example is the construction of multistorey buildings where the same trades (for example involving the sequence of plumber. ceiling fixer. Collectively. electrician. Other examples include pipeline and railway construction and estate housing projects. Many projects have the same sequential work repeating throughout the project lifetime.46 Example resource plots. and painter) are repeated floor by floor. Resource planning Resource planning refers to that involved in deciding the usage of resources on a project. and all other controls are held fixed. where emphasis is placed on the different level controls that can influence resource usage. The first method of handling resources mentioned above might be termed ‘parallel scheduling’.

because it can provide continuity in the workforce. This gives the total resource requirements on any day. Both resource smoothing and resource constrained scheduling practices use the resource plot as their starting point. For example. may be undesirable.47a). then this leads to simultaneous demand for resources. Peak demands of resources. for example. Smoothed resource numbers are desirable. if plant or equipment is the resource. Generally. Skilled labour is often difficult to obtain and expensive to fire. when activities are conducted simultaneously. resource constrained scheduling Problems associated with resource usage may be categorised as: • • Resource smoothing (also called levelling) Resource constrained scheduling.46 are project level representations. Smoothing the requirements for plant leads to higher utilisation and hence more efficient usage of these investments. Resource smoothing involves evening out the resource requirements over the project or over stages of the project. Some resource types can be expensive and limited in number (quantity). as scheduled in a bar chart.46. where emphasis is placed on the different level controls that can influence resource usage. and may be handled similarly. That is. and all other controls are held fixed. Resource constrained scheduling involves scheduling activities such that their resource requirements never exceed available numbers (quantity). The diagrams in Figure 3. eliminate undesirable hiring and firing and attendant industrial and people problems. Peaks in material requirements are to be avoided if only to avoid the possibility of shortfalls in supply during these peak time periods.Planning tools and terminology A–Z 103 The resource plot is developed for each resource type by summing up the daily (or whatever time unit is applicable) resource requirements for each activity. if people are the resource. then large numbers of equipment may not be available. producing peak resource demands at certain stages of the project. typical constraints will relate to upper limits on . by reducing peak demands for resource numbers (quantity) and creating a requirement for resources at other non-peak time periods. Materials problems may be similar to those for resources. Similarly it is not desirable to have plant and equipment being used intermittently (Figure 3. Plotting these values gives diagrams like Figure 3. particularly over short time periods. Resource smoothing. The resource smoothing problem and the resource constrained scheduling problem are subproblems of the total planning problem.

No further treatment of resources is necessary. Figure 3. people with certain skills or upper limits on numbers of pieces of equipment. A schedule is clearly not feasible if it requires resource numbers in excess of those available (Figure 3. When planning is carried out in an iterative analysis fashion. It is then necessary to carry out resource manipulation (resource smoothing or resource constrained scheduling).48). .48 Two broad approaches to dealing with resources. Resources are taken into consideration when establishing the nature and order of work.47b). two broad approaches may be observed when dealing with resources (Figure 3. Resources are ignored (or resources are assumed to be available) when establishing the nature and order of work.47 Schematic of what happens in (a) resource smoothing and (b) resource constrained scheduling.104 Conventional treatment from a systems perspective Resource Reschedule to remove peaks (a) Upper limit Time Resource Reschedule to remove constraint violation (b) Time Figure 3. Resource smoothing and resource constrained scheduling may or may not be significant problems depending on the degree to which resources are considered in the initial thinking.

an earthmoving exercise removes the hills and puts the earth into the valleys. rather than following any suggested order. The start dates of non-critical activities are moved. That is. but this usually comes with a penalty of increased cost and perhaps an additional and more concentrated resource requirement. Change resource numbers (quantity). then constituent level. incentive schemes. element level and project level approaches may be tried. These feed into higher level solutions. bonuses. for example. A control which is often tried first is: • Make use of any activity float.Planning tools and terminology A–Z 105 Most people seem to do something in between these two extremes. controls can be tried in any order the planner wishes. Note that shifting the start dates of critical activities will extend the project completion date. the solution steps to resource smoothing and resource constrained scheduling problems are no different to any project planning problem. then other activity level approaches may be tried: • Try compressing the durations of certain activities. These feed into higher level solutions. if the resource plot is thought of as hills and valleys. . The chosen work method itself can act like a constraint on the way resources are handled. if not always clear cut. Controls A common first approach attempting to solve resource-related problems (in an iterative analysis sense) is to use activity level controls. Solution (controls) at the element level. The solution could be expected to be more complicated when dealing with multiple resource types simultaneously. If activity level controls don’t work or don’t work completely. Shift work or overtime might be tried. Effectively. Compressing the durations of activities involves shortening the activity durations. they consider resources to a certain extent (maybe in their heads) up front in planning. resources and resource production rates) are selected to give the desired resource usage over the project duration. Some example controls follow: Solution (controls) at the constituent level. That is. Of course. carrots and sticks. Change resource productions through. In more fundamental terms. If this doesn’t work or doesn’t work completely. The solutions to both the resource smoothing and resource constrained scheduling problems are similar. the project controls (method. but also possibly fine tune later on. Solution (controls) at the activity level.

Within the solution approaches listed above. In place of more usual finish-to-start (F/S) relationships. This will lead to alternative network logic. though this may come with a cost penalty. sometimes called uneconomical drag out. starting the activity again after a break and so on. If no improvement occurs. Algorithms Many of the better computer packages that perform network analysis come with resource smoothing and resource constrained scheduling options. • Other activity level and project level approaches are only limited by the planner’s imagination. Examples include outsourcing or contracting out some work instead of doing the work in-house. using machinery instead of labour. and if it produces an improvement then more of the same is tried. The packages generally have inbuilt heuristic algorithms for solving these resource problems. try more of the same. If an improvement occurs. or other time unit). and prefabrication rather than in situ work.106 Conventional treatment from a systems perspective • • Try lengthening activities.49). The heuristic algorithms try something (such as the solution approaches listed above) (usually in increments of a day. Try something. Solution (controls) at the project level. Examine alternative methods of work. All activities then become non-critical. if no improvement is obtained. Try overlapping relationships between activities. Example project level controls are: • • Extend the project duration. . finish-to-finish (F/F). stopping the activity. then something else is tried (Figure 3. Knowledge of the project and the industry will help develop alternative approaches. Splitting implies starting an activity. there is still a need for the algorithm to prioritise what is tried. try something else.49 Heuristic logic of resource handling algorithms. or start-to-finish (S/F) relationships might be possible. start-to-start (S/S). Figure 3. Whether any of these approaches are possible will be determined by the project and the nature of the project work. Try splitting activities that can be split.

that performs resource smoothing and resource constrained scheduling. Rescheduling. Earliest finish time EFT (lowest or first occurring). Small projects can often have their resources rescheduled by inspection. rather some quantitative measure is needed for comparison purposes.Planning tools and terminology A–Z 107 If resource usage and activities are to be manipulated. . but a computer package cannot ‘eye’ the plots. leads to different resource plots. All heuristic algorithms require the planner to attach priorities and a ranking to activities with different attributes. Activity number (lowest). whether through such algorithms mentioned above or by hand. know that one resource plot is better than another? By human eye. The above suggestions are possibilities for establishing a heuristic algorithm for solving the resource smoothing and resource constrained scheduling problems. Computer solutions should also be checked for practicality. one plot can be seen to be better than another. Free float and total float (least or most). Such approaches will generally mean that a satisfactory solution may be found but one which is not necessarily optimal. Activities may be given rankings or priorities at the discretion of the planner (or computer package) according to any or all of the following as well as others: • • • • • • • • Earliest start time EST (lowest or first occurring). and so some form of solution that involves interaction between the planner and the computer package will perhaps lead to the better overall solutions. Measures How does a computer package. Duration (shortest or longest). Other suggestions may be found in the technical literature. Latest finish time LFT (lowest or first occurring). Where activities have equal priority then appeal is made to the next lower priority to distinguish between the activities. there is no definitive rule as to which activities should be manipulated and the order in which they should be manipulated. The allocation of resource numbers is carried out according to the selected priorities. Resource requirement (largest). Different resource plots may be compared quantitatively through the following measures: • • • • Sum of squares (moment) Number of hirings and firings Peak requirement Resource utilisation. Latest start time LST (lowest or first occurring). Each computer package could be expected to give a different solution.

For a resource requirement on day k of u k the sum of squares measure is given by SOS = all k uk2 (A sum of squares calculation is analogous to a statistical variance [standard deviation squared] [second moment] style calculation. a low number of hirings and firings. Compare the moment of inertia or second moment of area of a structural member’s cross section. the average requirements for two resource types might be 2 and 20 respectively. weighted measures can be used. the weightings would be expected to reflect this difference in magnitudes. for example an I-shaped cross section with that of a rectangular cross section. . For example.108 Conventional treatment from a systems perspective A resource plot with a low sum of squares. The weights take account of the relative magnitudes of the resource types.) The number of hirings and firings is the sum of the vertical portions of the resource plot. 1 = 1 2 Similar weighted measures can be devised for the other measures of hirings/firings. a low peak requirement and a high resource utilisation is preferred. 1 = 1 2 sum of squares for resource type i. The resource utilisation is the ratio of the area under the resource plot to the area under a rectangle drawn with one side as the peak resource requirement and the other side as the time axis. Weighted SOS = w1 SOS1 + w2 SOS2 + where wi SOSi weighting for resource type i. Compare the shape of a histogram or probability mass/density function with a high variance to one with a low variance. Multiple resource types When multiple resource types are present. resource utilisation and peak requirement. The peak requirement is the maximum resource requirement. The weightings are chosen subjectively to reflect the relative importance of the resource types. A weighted sum of squares measure would look like.

However. Some authors suggest that the weighting coefficients be chosen so as to give a minimum cost solution. not their absolute values. the measures (for example. The above use of weightings is an example of this. People resource and organisational (including coordination) requirements. and these are then traded off against each other. A solution is obtained for each resource measure separately. Such approaches are no more than is done in multicriteria (multiobjective) decision making or optimisation (Carmichael. The final solution is a compromise solution between what is best for each resource considered singly. The matrices are a way of informing project participants of their roles and duties. sum of squares) resulting from multiple resource types may be treated in several ways: • • • The measures for all the resource types are combined into a single measure. or similar. cost here being defined for each resource type as a unit resource cost (resource cost/day) plus a hiring/firing cost and the total cost being a sum over all resource types. One resource measure is regarded as more important than the others. and the less important measures are converted to constraints. Responsibility matrix Alternatively called responsibility chart. responsibilities and authorities can be represented in a responsibility matrix form which may have rows corresponding to the activities. The combination or combinations of weighting coefficients yielding the least cost solution is/are then used. For each activity. General In general. columns corresponding to resource types and entries corresponding to responsibilities. However. 2004). this cost is problem dependent and generalisations from one problem to another are difficult to make. . Hence to adopt such an approach for obtaining the weighting coefficients. a whole range of combinations of weighting coefficients has to be tried and the resource types smoothed and the cost evaluated for each case.Planning tools and terminology A–Z 109 The final resource plots depend on the values chosen for the weighting coefficients. Restricting the coefficients to integer values decreases the number of combinations that have to be searched. the number of combinations is still very large. there are attendant resource requirements. It is only the relative values of the weighting coefficients which are important.

. . review Uncertainty Criteria Objectives Figure 3. . source. alternatives Analysing risk situation Evaluation of consequences Selecting the best alternative Responding Figure 3.) Response. scale) Evaluation leads to Risks (high/low.50. Problem solving process Defining the problem Establishing the context Selecting objectives Risk management process Generating ideas. outcome (magnitude. yes/no. Potential loss/gain from event) Risk management The risk management process generally follows Figure 3. factor Analysis leads to Consequence.50 The risk management process. alternatives Identifying risk events Analysing ideas.51 Problem-solving process and risk management process compared (Carmichael. consequence Problem definition Event.110 Conventional treatment from a systems perspective Risk The exposure to the chance of occurrences of events adversely or favourably affecting the project/business/ as a consequence of uncertainty. Feeds back to source. 2004). . Risk = f (Uncertainty of event. uncertainty. treatment Monitoring.

for money. .Planning tools and terminology A–Z 111 The risk management process is essentially the same as that used in systematic problem solving. The basis for developing S curves is a bar chart. called a resource plot. (The cumulative plot is obtained from the resource plot or expenditure plot in the same way a cumulative distribution function is obtained from a probability mass function or probability density function in probability theory.52). The curve shape is like a stylised S for large projects (Figure 3. S curve An S curve is a cumulative representation of either a resource (people or equipment) or money over the project duration. The flatter portions correspond to the project starting up and the project winding down. going from project start to project end. histogram or profile.51. the S shape may not be readily apparent. with the section in between corresponding to ‘steady state’ progress of the project. Refer Carmichael (2004) and Figure 3. From the bar chart the resource (and money) requirements can be summed (integrated) over all activities on a day-by-day basis (or other applicable time unit). These daily figures (for each resource type. though different terminology might be used and the number of step classifications might be different. The S curve is an industry-accepted representation at the project level.52 Example S curve. where most resource and money usage occurs.) 100% Cumulative resource/money Winding down of project Project 'full steam ahead' Initial setting up of project Time Project start Project completion 0% Figure 3. called an expenditure plot or diagram) are then accumulated. to give the S curve (Figure 3. Those people familiar with soil sieve analysis will also recognise the steps of getting to the cumulative plot.53). For small projects.

. so a number of S curves are possible for any project (Figure 3.54) for each resource type and money. The S curve also finds use in reporting for replanning purposes.53 Example expenditure (upper diagram) and cumulative expenditure (lower diagram) plots. would generally be preferred and could be expected to lie somewhere between the two activity starting time extremes. For planning purposes. and may be used for budgeting purposes on second and subsequent similar projects. which leads to the more agreeable resource and financial commitments. for example. as opposed to an expenditure plot or diagram that is equivalent to a resource plot in presentation. The S curve for money may be referred to as a cumulative expenditure plot or cumulative cost plot. It is an industry-accepted way of representing planned expenditure at the project level. Because the bar chart is not a unique entity for any project. the particular S curve. some activities have float and can be adjusted in their start times (between earliest start times EST and latest start times LST). Cumulative expenditure plots are repeatable to a reasonable accuracy between similar projects.112 Conventional treatment from a systems perspective Expenditure $ 0 0 Days 100% Cumulative expenditure 0% Project start Project end Figure 3.

the extent of the work to be undertaken. is usually referred to as scheduling or programming. Scope Scope is what is involved in undertaking the project. what work is contained in the project (perhaps by defining what work is not included in the project. that is when work is to be done (and hence it follows. The biggest transgression though is the sloppy.Planning tools and terminology A–Z 113 100% EST S curve Percentage of resource or money Envelope for possible S curves Final planned usage LST S curve 0% 0% Percentage of project time 100% Figure 3. Some writers wrongly include extra things besides project work. in effect the boundaries to the project) that leads to the end-product. some people enlarge upon this meaning and loosely use the term interchangeably with planning.54 Envelope of S curves based on activity start times. Commonly the carriage is put before the . ‘constraint’ and ‘scope’. few people appreciate the distinction and why there is a distinction. There is much confusion over the usage of the term ‘scope’. Scope is fully described by listing all the activities. Scheduling Establishing the timing and order of work. interchangeable use of the terms ‘objective’. and should be clearly distinguished from scope. it is not scope. Schedule Refer ‘program’. when resources and money are needed). While useful information to have. However.

State is a controlled variable. and including program management) is not favoured in this book. Neither form actually manages time. some people loosely refer to planning and time management synonymously. such that the scope is said to determine the objectives and the constraints. it is a vector quantity of state variables). Time-scaled network A time-scaled network is an activity-on-link network drawn to a horizontal timescale. Time-chainage chart Refer ‘cumulative production plot’.114 Conventional treatment from a systems perspective horse. Refer Carmichael (2004). Task Refer ‘activity’. that is there are no observability (in the sense of Kalman) issues. Time management The term ‘project time management’ (that is. but the terminology is widespread. There is no vertical scale. Stage A stage corresponds to a subinterval of the project duration. The lengths of the activities (links) are made proportional to the activities’ durations. The term ‘time’ is used correctly when used in a technical sense. It may have physical interpretation or not. but is used in a lay person’s sense elsewhere. the management of anything involving time on a project. For the stripped-back projects considered in this book. There also exists the term ‘personal time management’ relating to how people ‘use their time’. A time-scaled network . output and state are the same. time is viewed as the independent variable. State contains multiple parts (that is. In this book. State The state describes internal system behaviour. Also. and projects are regarded as dynamic systems because of this. This confusion and imprecision stifles the understanding of project management.

In time-scaled networks. Float is usually indicated by lines with texture different to that used for the scheduled duration of activities. different drafting styles are encountered.1 and networks. base cpts Approval. Planners use whichever style appeals to them.25(a. the vertical lines in Figure 3. public holidays. how long activities take.55 might be drawn inclined. Figure 3. base cpts Deliver base cpts Manuf.55 Time-scaled network for an example project. See the associated bar chart. (Activity names have been abbreviated because of space limitations – work items are implied.56. As with networks and bar charts. For example. Activity-onnode networks do not lend themselves directly to the time-scaled version.55 gives an example time-scaled network.Planning tools and terminology A–Z 115 gives the same information as a connected bar chart. dsgn. number of shifts per day and so on. and the dependence between activities.) . weekends. the amount of float in activities. Figure 3. rostered days off (RDO). namely when activities begin. options Marketing Assemble base cpts Apply finishings Assemble options Approval. The information from a network analysis is readily transferable into a timescaled network form. A time-scaled network may be referred to as a program or a schedule. options Sales Scheduled duration Float or dummy activity Figure 3.b). design Det. dsgn. A time-scaled network follows directly from a network analysis. when activities end. the vertical components have no consequence. 0 10 20 30 40 50 60 Days 80 Prelim. However this does not stop some people from pretending to draw activityon-node diagrams to a timescale. The horizontal scale is ideally matched to a calendar allowing for stipulated work hours per day. Only activity-on-link networks can be drawn to a timescale. Figure 3. base cpts Det. Typically such diagrams end up looking like Figure 3. activity durations are given by the horizontal components of lines.

it finds application at all phases of a project from concept through to termination. A true time-scaled activity-on-node diagram would have elongated and contracted nodes (circles. diagrams like Figure 3. Then it was a design review or ‘second look’ approach to proposed or existing designs. Value management/analysis/engineering Value management (equivalently value analysis. It identifies . As such it can be shown to be but a special case of systems engineering/problem-solving methodology. Instead.56 Pretend time-scaled activity-on-node diagram. Its area of application has enlarged over the intervening years such that it now encompasses not only design reviews but also. Time-scaled networks also find use in reporting for replanning purposes. activity-on-node diagrams over activity-onlink diagrams. feasibility studies. Constructability or buildability studies are but special cases of value management applied to construction or building projects.56 are drawn. when a planner uses a connected bar chart or a time-scaled network. rectangles ). Its origins trace back to about the 1950s when it was known as value analysis or value engineering. the planner is unwittingly using an activity-on-link diagram representation. value engineering) is popular in project management for a diverse range of situations. Originally conceived as a way of economising on resources or finding alternatives. and would look silly. or show a biased preference for. for example. Timetable A timetable is a document showing when events occur. and conflict situations.116 Conventional treatment from a systems perspective Activity A Duration 10 Activity B Duration 5 Figure 3. Central to value management is the analysis of function from a whole system viewpoint and the proposing or generating of alternatives. So. even though that planner might confess to being a staunch advocate of. Central to the approach is a systematic analysis of function and an examination of alternatives. but these are activity-on-link diagrams disguised to look like activity-on-node diagrams. It may be referred to as a program or a schedule. an examination of project goals.

and can assist in testing assumptions and needs. It is not referring to any probabilistic measure of variability. Variance A variance is the difference between planned (or as-planned) and actual (or as-executed. Refer Carmichael (2004). However. The whole system or holistic approach avoids conventional compartmentalised thinking and obtaining locally optimal or suboptimal solutions at the expense of the desirable globally optimum solution. There. as-built. under budget. and not the end-product. as-made. what is happening is more akin to generally defined processes being broken down into more refined processes. Some writers (wrongly) say that scope delineation involves subdividing the project deliverables. Work breakdown A project (scope) can be decomposed into subprojects and then into subsubprojects and so on down to the activity level. Work breakdown is not unique. A breakdown is chosen that makes life as easy as possible for the planner. special definitions of variance are used. . A variance is positive if the project is ahead of production/schedule or. The process of decomposing a project (scope) might be called scope delineation (or scope definition) by some writers. The degree of coarseness or fineness with which a project is subdivided will depend on the end use of the plan. Variance = Planned − Actual The usual convention is for a variance to be negative if a project is behind production/schedule or. (Deliverables are the end-product plus project state final conditions. as-constructed. including cost variance and production/schedule variance. for example. over budget. or as a product-oriented tree. only better or worse subdivisions. There is no right or wrong answer to this. The lower levels represent finer detail. duplication and unnecessary expenditure.Planning tools and terminology A–Z 117 wastage. ). Some people (wrongly) see this breaking down as going from the ‘product to the processes’. Using earned value. for example. since a project is not a ‘product’. The term work breakdown structure (WBS) is used to refer to the resulting systematic decomposition (as opposed to an unstructured activity list where there is a possibility that some activities have not been thought of). various work breakdowns are possible for any project.) But only project work should be involved in the breakdown. several types of variances can be identified.

Besides the financial incentives. critically examining the facts and the sequence and developing alternatives. It is commonly represented by a network. Work study involves the critical and systematic analysis of work with an ultimate view to improvement and the elimination of any nonproductive components. Method study breaks down work into its components and questions the purpose and need of each component. process or activity. Work method Work method is the sequence or logic of operations. subproject.118 Conventional treatment from a systems perspective Planners distinguish between events and activities. Terminology for each of the levels. equipment and materials. Each activity has events associated with it. There is no consensus in the use of terminology to describe the various levels. used by different people. Work study is an established technique that has wide applicability. Work study is made up of method study and work measurement. recording information on the work. Some writers (wrongly) include events in work breakdown structures. Planning with either is okay. but care has to be exercised when the two are mixed. might include project. equipment and materials in work tasks. it attempts to create an attitude of mind about the effective use of people. Work study Work study attempts to create in people a questioning state of mind in the way they view their current and future work practices. work package. Re-engineering is work study reinvented (Carmichael. 2004). amongst other things. This aware state of mind applies to all phases of a project and on a continuing basis in an organisation. ‘Program’ is a term used to refer to the level above projects and represents a collection of projects. . Work study may be broadly defined as an examination of the use of people. Work measurement is concerned with the time periods taken to perform work. Work study can be shown to be a special case of systems engineering/ problem-solving methodology. An alternative usage of the term ‘program’ is given above. job. though this book’s preference is for the terminology project – subproject – – activity. it involves. and an associated attempt at improvement and elimination of waste. that is a project is a subsystem of a ‘program’. subjob. with activity and event interrelationships. The terminology of task and subtask might also be used at the middle to lower levels.

contract/procurement management. or based on invoiced. Under what circumstances should incomes and expenditures. and (iii) resources. What do you think of such a practice? Some planners will include delays in all programs except at the very detailed level. or other? 2. By including delays does it take away the emphasis on target dates? Will knowledge that an allowance for delays has been included in the program make people think that delays are inevitable or expected? Is this a case of Parkinson’s Law? – Work expands so as to fill the time available for its completion. organising. ‘controlling’ and coordinating that imply involvement with (i) schedule/production. human resources management and communication management that imply involvement with (i) schedule/production.17. at the lowest level it is believed that by incorporating delays. directing. be plotted based on dates of anticipated transfer of funds and billings. The argument behind this is that the overall program should present as realistic as possible picture of the duration of the project. will designers take longer knowing that a delay has been incorporated into the site work program? Should delays only be incorporated in the broad program. quality management. (ii) money and (iii) resources. staffing. committed or paid amounts. industrial action or other unforeseen circumstances. (ii) money. risk management. Some planners include in their program an allowance for delays – due to weather. ‘time’ management. cost management. as shown in Figure 3. These delays are incorporated into programs as separate activities. How might this ‘classical’ management function division help or hinder the understanding of planning as outlined in this book? (b) Isolate the parts of the management function approach of scope management. Is this a case which supports the notion of self-fulfilling prophecies? Will some of the project team assume they have longer to do their work because a delay has been included in the program for other members of the project team – for example. made public) and the other with delays (kept private)? . with detailed programs omitting delays in order to focus attention on target dates? Should there be two programs – one without anticipated delays (a target plan. focus on target dates disappears. (a) Isolate the parts of the ‘classical’ management function approach of what are categorised as ‘planning’.Planning tools and terminology A–Z 119 Exercises 1. How might this management function division help or hinder the understanding of planning? 3.

120 Conventional treatment from a systems perspective

4. Obtain a user’s manual for a computer program that does network analysis including resource handling and overlapping relationships. Research the program input and program reporting facilities. 5. For the following project network in activity-on-link diagram form, convert it to an activity-on-node diagram form. For convenience, activities have been listed as letters of the alphabet.

A

D F

C E B

G

Convert the following project network in activity-on-node diagram form to an activity-on-link diagram form.

B

C

F

A

D

G

E

H

6. Why might some planners dogmatically criticise activity-on-link diagrams (while stating a preference for activity-on-node diagrams), yet quite happily use bar charts? 7. An S curve (money) represents cumulative expenditure. Why is this form of diagram preferred by industry over reporting on individual activity expenditure, much like a bar chart reports on individual activity durations? 8. In examining some proposed forestry (timber harvesting) operations, an environmental impact statement is prepared. The main activities in this are listed in Table E3.1. The order in which the work is planned to be done is according to the network logic of Figure E3.1.

Planning tools and terminology A–Z 121 Table E3.1 Forestry study activities Activity code Object Descrip Altern Physic Atmos Hydrol Veget Fauna Econo Cultur Coord Report Activity Establish proposal objectives and need Develop description of proposal Examine alternatives Study physical environment and the effects of forestry activity Study atmospheric environment and potential impacts Study hydrological environment and impacts of the proposal Study vegetation and effects of the forestry activity Study the fauna and effects of the forestry activity Examine economics and land-use effects Examine cultural environment and effects of the forestry activity Coordinate project Produce report Normal duration (days) 10 3 2 20 6 15 20 20 15 10 25 55

Report Object Descrip Altern Physic Hydrol Fauna Econo Atmos Veget Cultur Coord

Figure E3.1 Forestry study logic.

Assume each of the ‘Study’ activities may be reduced in duration by 20% and the remaining activities by 10% (each rounded up to the nearest day). What is the possible shortest duration?

Chapter 4

The planning process

Introduction
Planning dimensions Planning starts with broad assumptions which are steadily refined, as more information comes to hand. Planning typically proceeds from the broad to the detailed. Planning is typically done from the general to the particular, from coarse to fine, in two directions (or two dimensions) (Figure 4.1) at the same time: • In hierarchical levels involving breaking (the work in) projects into subprojects (for subprojects other than that based on stages), subsubprojects, , activities, elements and constituents. Detail emerges with lower levels. Project work is dissected into manageable portions through work breakdown (discussed later). Over time. Detail evolves as the project progresses.

The launching point for the main project planning is at the time of establishing the initial scope (Figure 4.1), which derives from the objectives and constraints. For each and all levels and at any time in Figure 4.1, a planning problem exists. A plan near the top left of Figure 4.1 might be called a ‘project preliminary plan’ or similar name. It contains broad information on work method, resources, dates and so on. And it may be used as a basis for approval to do further project work, as might other plans at later stages.

As a synthesis problem Planning is a synthesis problem. As such, there are multiple solutions (choices of control) possible. In most cases, planners are only after a satisfactory solution, or a solution that they can live with, and do not put

The planning process 123

Detail evolves over time Coarse Detail unfolds with project, subproject, . . . , activity, element and constituent Coarse Initial scope Fine Final scope Stage

Fine

Level

Figure 4.1 Evolution of planning problems.

additional effort into searching for the optimal solution. A planner may also be under pressure to come up with quick solutions. Planners are unaware of and do not understand the components of the synthesis problem, and so they never know where they are relative to the optimum. They are unable to vocalise or formulate the synthesis problem components. Such discussion goes to the very heart of current planning knowledge being in its infancy, and current planning education (read training) being superficial and low standard. As an iterative analysis problem The solution of any planning problem via an iterative analysis mode involves the conventional steps in systematic problem solving or systems engineering. Feedback occurs between steps in trying to refine the problem. (This is additional to the feedback that can occur between levels.) Alternatives might be arrived at through the generation techniques used in a value management/analysis/engineering study, which aims at developing alternatives that perform a desired function but at a lower cost, or in a work study, which is concerned with the method and timing of work activities and their optimisation. Both value management and work study are no more than versions of systematic problem solving (Carmichael, 2004).

Case example – Small commercial buildings This case example looks at one contractor’s procedures for planning the construction of small commercial buildings involving minor

124 Conventional treatment from a systems perspective

earthworks, concrete work, stormwater drainage, steel fabrication/erection and electrical fitout. Three phases/types of planning – initial, detailed and ongoing – are used. The initial or preliminary planning occurs during preparation of a tender. A program is developed for submission with the tender and to assist the estimating department in pricing the project; the program may be required to assist the owner to further assess the contractor’s offer. The project duration may be set by the owner and it is a case of allocating sufficient duration to each activity so that the overall project is completed within the nominated duration. The detailed planning does not occur until the tender has been successful. When a contract has been awarded, and within the first couple of weeks, a detailed construction program is developed. It is commonly a contract requirement that a construction program be submitted within 14 or 21 or 28 days of contract commencement. Work on the construction program incorporates a review of the initial program, specifications and drawings. After possible discussions with the site foreman, it is determined which activities of the initial program can be retained and which activities require further breaking down. The aim is to develop a program with sufficient detail that can be easily followed and updated. Experience helps in determining activity durations and overlapping relationships to complete the project in the nominated time frame. This information is then input to a commercial project management computer package so that the critical paths, floats etc. can be determined and so as to have the program in a professional form for submission. The program is presented as a connected bar chart. Ongoing planning of the project leads to updating the construction program on a monthly basis and the preparation of fortnightly programs (programs for the following two weeks). The fortnightly programs contain considerably more detail than the construction program and are usually updated on a weekly basis.

Planning problem components
The main components of any (optimal) synthesis problem (Figure 4.2), to which planning belongs, are (Carmichael, 1981): • • • Model Objective(s) Constraints.

The planning process 125

Optimal synthesis (inc. planning) problem components

Model

Objective(s)

Constraints

Figure 4.2 Optimal synthesis and planning problem components.

Objectives and constraints are developed below. Models are developed separately (Figure 4.3): • • • Hierarchical models – Chapter 5, Chapter 9 Multistage models – Chapter 8. The project is broken down into stages, or equivalently the project duration is discretised Probabilistic models – Chapter 7.

These problem components are expressed in terms of control, state and output variables. The system notions of control, state and output variables are central to the development of models and thinking for planning purposes.

Models

Hierarchical models

Multistage models

Probabilistic models

Figure 4.3 Model development in this book.

126 Conventional treatment from a systems perspective

Control Control is equivalent to a planning decision, action or input. The term ‘control’ is favoured here in place of decision, action and input. The planner exerts control on the project in order to bring about a desired project behaviour. The most obvious controls are work method, resource (people and equipment) selections and resource production rates. Work method, resource selections and resource production rates may be freely selected by the planner, subject to any constraints being present. The preferred controls are those that extremise the objectives. State, output The output describes external or observable response, performance or behaviour. Internal behaviour is usually given in terms of system state. By the nature of projects (in the stripped-back form treated in this book), the internal and external behaviour are the same. (The observation equation of control systems theory [Carmichael, 1981] becomes ‘output = state’.) There is a one-to-one transformation between output and state. The terms ‘output’ and ‘state’ become interchangeable when referring to projects, and there is no noise in the observation device. There are no observability (or controllability) (in the sense of Kalman) issues for projects thought of in a stripped-back form. The most obvious states are the resources used so far (cumulative resources) or the cost so far (cumulative cost) and the production so far (cumulative production). Disturbance In most systems studies, there is something that prevents the hoped-for output being obtained exactly. Commonly this is because of a changing project environment brought about because of delays, disruptions, adverse climatic effects and so on. Collectively the corruption is referred to as disturbance below, though control texts may use the term ‘noise’ (Figure 4.4).

Disturbance

Control/input

(System) State

Output

Figure 4.4 Schematic representation of control, state and disturbance.

(After Carmichael. the planner settles on one particular sequence of work. when money is used. in what order and when and with what resources (additionally expressed in a money unit). 2004) Simplifications to planning Ideally. alternative cash flows and project durations evolve. and at what resource production rates. This establishes when resources are used. duration of the work and resource availability. requiring ingredients of creativity and . connected bar chart Constraints – Related to work duration. Using an optimal control approach. There may be constraints on costs. so resource utilisations. The full planning problem and its solution are very complicated. However. As such. the tools are not currently available for this as a stand-alone mathematical problem. and thereby the duration of the work. A suitable system model is the time-scaled critical path network or connected bar chart. Ultimately. the controls available to the planner are the order or sequence of the work including when particular work items will be done. Objective(s) – (Minimum) cost and/or (minimum) duration. costs.The planning process 127 Example Planning establishes how and what work will be carried out. if expressed in a monetary unit. Consider the simpler (subsystem) problem where it has already been largely established how the work will be done and with what type and number of resources. with development to resource plots and cashflow diagram. or. resource availability and resource production rates. the components to the planning problem are: Model – Time-scaled network. By a series of iterations. project planning should be carried out as a probabilistic (inverse) synthesis problem. what the planner might be trying to achieve is a work sequence of minimum cost or minimum duration or a compromise between these two objectives. What the planner is trying to influence is the utilisation of the resources (state variables). As the planner adjusts the sequence of work.

Analysis is much easier than synthesis.5 Simplifications to make the planning problem more tractable. However. Instead. In some cases. then there may be only one iteration (or even zero iterations) needed to give a sufficiently good answer. Subprojects may be chosen such that the interaction between subprojects is minimal and is commonly neglected. Iterations may modify the method. the step of generating alternatives is relatively straightforward without knowing how close to the optimum the guessed solution is.128 Conventional treatment from a systems perspective Planning problem simplifications Conversion from synthesis to iterative analysis Hierarchical breakdown of a project Conversion from probabilistic to deterministic Figure 4. planning is simplified and made more mechanical in three main pragmatic ways as shown in Figure 4. The levels are uncoupled as far as possible. elements and constituents. productions. experience. resource usage and resource production rates until satisfactory (without necessarily being optimum). • Instead of solving the planning problem in a synthesis format. Where future events have • • . and so on to activities. The project (work) is broken down (hierarchically) to lower level subprojects. durations. where the planner is feeling his/her way. interaction at levels lower than subproject is commonly strong. costs are made deterministic in nature. to make the level problems less complicated. Estimates for resources. The analysis is made deterministic. permitting subprojects to be planned independently of other subprojects. it is converted into an iterative analysis format. sub-subprojects.5. The same holds true where multiple projects are combined on an organisation level. Iterations (between levels) can largely be removed by the planner being aware of what is happening on other levels when solving an individual level problem. Interaction occurs because of the work method (implying an order/sequencing of work). and if a good initial guess is made. for tractability reasons. Future events (leading to both upsides and downsides) are anticipated. the analysis may become a ‘what if’ or sensitivity analysis examining the impact of alternative controls. The planning problem is assumed deterministic. For most planners.

Planning stage by stage. may be used. to help in the generation of alternative controls for replanning. Replanning Replanning is carried out according to stages or at discrete points in time. and other planning matters.The planning process 129 a likelihood (uncertainty). the ensuing scope and planning is ill-defined. However. According to Machiavelli in The Prince: When trouble is sensed well in advance it can easily be remedied. for example for an owner at project outset. there some probabilistic analysis technique. Information assisting the planning problem could be expected to be better the less far into the future. that is. but this follows from the nature of what replanning is. Replanning is based on the remaining duration of the project using the current state of the project as initial conditions. . A good plan allows for some flexibility and can be adapted to changing circumstances that may occur during a project. and comparing this with the current baseline (‘as-planned’). attempts are made to foresee potential influences in upcoming stages. Whether replanning is necessary or not is established through monitoring/sensing the project output (‘as-executed’) through time. Scope is selected based on stated objectives and constraints. without consideration of upcoming stages. Planning at some stages may take on increased importance for different stakeholders. if you wait for it to show itself any medicine will be too late because the disease will have become incurable. Replanning is no different in form to planning. Objectives and constraints General Objectives and constraints derive from the value system of the project/ end-product owner. Objectives and constraints lead to scope. Without defined objectives and constraints. this is incorporated into thinking akin to risk management. The planning problem (for the remainder of the project) becomes better defined as the project progresses. such as Monte Carlo simulation. The difference between the ‘asplanned’ and ‘as-executed’ values (together with an updated view of future influences) may then be used to indicate what controls need to be modified. and for a contractor when tendering and while undertaking the work. and controls are chosen with these influences in mind. but more usually is handled deterministically through a ‘what if’ or sensitivity analysis. Planning is always carried out over the full project time span. could be expected to lead to a suboptimal solution.

but its intent is not in a systematic problem-solving sense. And these may apply throughout the project (for example minimum deviation from specification). element and constituent) and over time. The criteria by which the preferred materialisation of the end-product is selected are the project objectives.6 End-product and project objectives. The materialisation of the end-product can be performed. That is. possibly. and hence there is no unique solution.130 Conventional treatment from a systems perspective The term ‘planning objectives’ can occasionally be found in writings on planning. general project objectives will contain a component over the time domain of the project and a component at the final (terminal) point. activity. or at the final (terminal) point of the project (for example minimum [duration which is related to the project] completion time). and later times. asset. . They may be decomposed to finer detail and stated more specifically for use in lower levels.6. project [duration] and deviation from specification. but other objectives are possible. Form. there are possibly an infinite number of versions of end-products. In a similar idea. the problem is an inverse problem. Scope and other planning considerations on projects follow. In systems studies terms. finishes etc of the end-product follow. and hence is not used here. service etc. Objectives End-product objectives Project objectives Figure 4. Objectives There will be an identified need or want for some product. The selection of the preferred end-product form is a design problem. This end-product is achieved through a project. The objectives and constraints carry through to all planning problems at all levels (project. Commonly. function. facility. on any project there are two types of objectives • • End-product objectives Project objectives. in an infinite number of ways. project objectives say something about project cost. The criteria by which the preferred end-product is selected are the end-product objectives. That is. See Figure 4.

Where there are multiple objectives. net present value NPV project – cost. For any situation there may only be a single objective or there may be many objectives (multi-objective). 2004). for example. The higher levels and early project stages constrain the choices in the lower levels and later project stages respectively. budget ) Duration (end-product – lifetime . to: • • • • • • • • • • • • Money (end-product – sales. Detailed planning is established within the parameters of previous broader planning. The solution of such multi-objective problems generally involves some subjectivity while the solution of the single objective problem does not. they restrict the options that are possible. Confusion can arise when people start expressing project objectives in terms of the end-product requirements or end-product objectives. as the case may be.The planning process 131 Example issues included in objectives End-product objectives and project objectives (and constraints) may relate. benefit:cost ratio (BCR). These are expressed in terms of end-product matters or project matters. Constraints All projects have genuine constraints such as funding. many people confuse a project constraint with an end-product constraint. environmental (natural) and political. One or more of the objectives may be more dominant than the others. As with objectives. End-product constraints may influence project constraints (Carmichael. Constraints limit the range of controls possible. they may be non-commensurate and conflicting. cyclones ) Social impacts Geotechnical considerations. project – duration ) Resource usage Quality issues Community acceptance Environmental (natural) effects Safety Risks Public impact Extreme event impact (floods. .

Cost. environmental considerations Allowable sound volumes Site space restrictions. Profitability. Example objectives (project. the preferred planning solution will lie on either the duration constraint or the cost constraint. capabilities Transport. . Project completion time (may be imposed or as-planned) Production/schedule delays due to. That is. legal requirements. Cash flow Hazards and local factors. in the required quantities when needed A certain specified amount of interaction with the community Complete the work for less than (money). access. site access limitations. element and/or constituent levels) • • • • (Minimum) impact of activities on the natural environment (Minimum) project duration (Maximum) utilisation of resources (Maximum) return on money invested (in project and endproduct) – related to usage of available money. storage. of the required type and in the required quantities when needed. Spending limitations. element and/or constituent levels) • • • • • • • • • • • • • • Availability of resources. It is frequently the case that a duration constraint or a cost constraint will be binding in the planning solution. funding approvals. activity. quality assurance Designated milestones.132 Conventional treatment from a systems perspective Constraints can be converted to objectives (and vice versa). a smoothed requirement for resources Availability of money. Deadlines. safety issues Statutory. Example constraints (project. activity. work flow A lower limit on the standard of work. Budget. manpower availability problems and weather problems Events that are required to happen at the right time Technical knowledge. delivery issues. for example. resource movement. regulatory approvals.

and the owner instructed the company to proceed and commence design. The time period before the construction started then became the time period available for design and documentation. The hours given by the sections formed the basis of the fee proposal presented to the owner. a program was developed to show how the work had to be done in order to meet the end date. This resulted in a six-week time period in which to design and document the job. A certain desired standard/quality in performance may leave little flexibility in the choice of resources. does not prevent the free choice of sensible controls (method. Case example – Fruit processor upgrade This case example discusses the practices used within a company to assign resources to a project involving an upgrade of an owner’s existing fruit processing. based on the scope (of work) for their sections. Once the project manager had determined the resources required on the project. and the completion date was fixed by the need to have the work completed and the fruit processor running when the first fruit was ready for picking. The upgrade involved civil. after allowing a sufficient time period for tendering. this may represent. and were asked to advise who they would assign to the job. the section heads advised the number of hours they required to complete their work. Hence if the project proceeds in whatever form. structural. section heads assigned these resources to the project. a financial loss or a completion time overrun. mechanical and electrical work. or the scope may have to change – ‘you get what you pay for’. . In the extreme case. The company doing the project work was structured in three sections along discipline lines – civil/structural. especially an unrealistic cost constraint. process/mechanical and electrical/instrumentation. the constraint will be violated. a constraint imposed without thought may be such that there are no feasible controls. resources and resource production rates). for example. all the documentation had to be completed before the project manager went to site. At this stage. They were given an indication of the program.The planning process 133 Care has to be exercised that any imposed constraint. The other factor driving the design phase program was that the project manager was going to site for the construction phase and. The construction program was developed backwards based on the defined end date and the delivery dates for some long lead time period items that had been previously ordered. in order to most efficiently run the design phase. Following scope delineation.

134 Conventional treatment from a systems perspective

At this stage the project manager felt that all necessary resources were in place because all section heads had the hours they had asked for, and the project was proceeding on schedule. As the design phase proceeded, it was found that the people who had been identified to work on the project were being diverted from this project onto other projects, and the design began to fall behind schedule. Discussions with the section heads brought no solution and it was only after the project manager complained to one of the company directors that pressure was brought to bear on the section heads and the project was assigned the resources it needed. At this point the design was about two weeks behind schedule. Fortunately, it was possible to get the critical documentation done during the period of reduced resource availability so that the construction schedule did not suffer greatly. Off-setting this partly was a delay in receiving some information from the owner, which also delayed the documentation for a component of the work, and so the lack of resources available for this particular work package was not a problem. The project encountered difficulties because it was only one of a large number of projects which were then happening, and which required resources from a finite pool. Scheduling the use of resources could be difficult at the best of times, because owner approval for projects was often delayed, sometimes for weeks, and the company had to try and juggle the resources it had among whatever projects were going at any one time, and hope that the projects lying in the wings did not all happen at once. In this case a number of other projects, that had had owner approval delayed, all got the go ahead just as this project was starting. In addition, a number of new projects ‘came in the door’, and hence there was a large demand for resources that the company did have. The section heads did not carry out any formal resource smoothing across all the projects requiring resources. Resources seemed to be assigned on a perceived priority and on a ‘whoever screamed the loudest’ basis.

Scope
Scope is what is involved in undertaking the project, the extent of the work to be undertaken, what work is contained in the project (perhaps by defining what work is not included in the project, in effect the boundaries to the project) that leads to the end-product. Scope is fully described by listing all the activities.

The planning process 135

There is much confusion over the usage of the term ‘scope’. There is much loose usage of the term ‘scope’. The biggest transgression is the sloppy, interchangeable use of the terms ‘objective’, ‘constraint’ and ‘scope’; few people appreciate the distinction and why there is a distinction. Commonly the carriage is put before the horse, such that the scope is said to determine the objectives and the constraints. Getting to the scope is the process of working through the inverse problem associated with the ‘means to the end-product’. Objectives and constraints are delineated, alternative work methods are proposed and evaluated against these objectives and constraints, and a best work method selected. This defines the extent of the work or scope. As a project progresses, the objectives may change but more likely the constraints will change (for example, the work completed so far constrains what can be done next, assumptions on delivery time periods were inaccurate ). This leads to scope changes. The extent of the project work follows from a knowledge of the project objectives and constraints. Work established without reference to objectives or constraints could be expected to be non-optimal with respect to the objectives, and also possibly violate the constraints. Some people are able to rationalise this as being satisfactory, and some even rationalise it on the basis that they are being ‘practical’. In reality, though, such people are unaware of how the scope follows from a knowledge of the project objectives and constraints. Attention is paid to scope in order to indirectly [contain] deviations in cost, [schedule] and quality from that planned. Scope changes on projects are common and may be referred to as variations or extras. Some projects have a tendency for multiple changes leading to what is sometimes termed scope creep. Scope changes may come about through changed thinking, errors, changed values and circumstances, unforseen issues and so on. People confuse a change in the end-product with a change in scope; the first will commonly lead to the second, but they are not the same. People also wrongly talk of control when they speak of ‘change control’; invariably what is involved is a monitoring and recording of changes and perhaps some restrictions on changes. Changes then lead to replanning of the remaining part of the project. Scope becomes better delineated as the project progresses. For many projects it is not possible to know the exact scope at the start of a project. (Carmichael, 2004) From objectives and constraints to scope Scope derives from the project objectives and constraints. Figure 4.7 shows the steps involved. (Planning concerns itself with the ‘means to end-product’

136 Conventional treatment from a systems perspective

End-product

Means to end-product

Identify need for end-product

Creates need for means

Establish objectives and constraints

Establish objectives and constraints, deriving from: (i) End-product objectives and constraints (ii) Elsewhere

Generate alternatives

Generate alternatives

Feasibility/select preferred end-product

Feasibility/select preferred means

End-product concept

Broad scope

Design development

Plan development

Figure 4.7 Broad design and planning steps.

problem. The parallel problem of design concerns itself with the ‘endproduct’ problem.) The means to the end-product is based on knowledge and experience relevant to that project’s industry, and an understanding of that industry’s projects. Ideally, all available ideas are evaluated. Viable alternatives/options are assessed including the carrying out of cost–duration trade-offs. Peculiarities of the project are considered.

Scope and planning Scope establishment is part of the initial broad planning problem. In Figure 4.1, this is referring to the top left corner.

The planning process 137

Extras feeding into and out of planning
Into For planning to take place, supporting information is needed to feed into the basic problem-solving steps. Supporting information includes (Figure 4.8): • • Assembled and identified project data, for example, that obtained through a site inspection. Some familiarisation of the project environment and the nature of the work commonly takes place. Estimates of activity durations, associated resources (people and equipment), resource production rates and money to carry out these activities – availability of resources and money in time (when) and quantity (what). Estimates can be refined as the project progresses, and more information comes to hand; this in turn can lead to more accurate and more detailed plans. The behaviour of people (Chapter 7) including people’s motivation, resistance to change and work ethos. Potential difficulties are anticipated. Forecasts are made and assumptions behind forecasts are tested.

• •

Communication The outcome of planning, together with the underlying planning assumptions, are communicated to the project stakeholders. Various forms of communication are possible, usually as documents (but can include physical and computer models) and verbal communication. The documents provide a basic reference and briefing for those who will ultimately be involved in the project execution. The documents allow project outsiders (including senior management) to review the plan, provide guidance and assist and, thereafter, support the plan throughout project execution. If, following review, the plan is found to be satisfactory, the plan is implemented.

Project data, Estimates, Availabilities, People behaviour, ...

Planning

Programs, Communication of the plan and underlying assumptions, ...

Figure 4.8 Extras feeding into and out of planning.

138 Conventional treatment from a systems perspective Table 4.1 Commonly used graphical forms of communication Information with respect to Schedule Resources (people and equipment) Money Production, work done; activity/project production rates; material usage Delays Communication means Bar chart, time-scaled network Resource plot, list; cumulative resource plot (S curve) Cumulative money plot (S curve), cash flow diagram, list Cumulative production plot Cumulative delay plot

There are a number of useful ways to assist the planner organise and present planning information. These take the form of charts, plots, activity date/time listings and so on. They might be regarded as the end-result of the planning process to date. Commonly used graphical forms of communication are indicated in Table 4.1 (refer Chapters 3 and 6 for descriptions of these). As communication tools, there is a requirement for clarity, explicitness, being uncluttered, without superfluous information, and ease of understanding (intelligible) for all the end-users. Covering reports can be used to state assumptions, calculations and so on, and to keep the diagrams uncluttered. Colour can be used to improve interpretation. According to Murphy’s Law (extended): The most important piece of information in any plan or document stands the greatest chance of being left out.

Programs
General Programs are a means of communicating the outcomes of planning. Ideally, some comment on the underlying planning assumptions accompanies the program, in order that end-users are well-informed, and can see the bigger picture. The forms programs take reflect what is trying to be communicated. Programs may be any of the following: • • • • • (Connected) bar chart Time-scaled network diagram Cumulative production plot Activity date/time listings/timetable Marked-up drawings.

The planning process 139

They ideally derive from a critical path analysis. They may commonly be presented: • • Over different project levels Over different time horizons.

Associated with such programs, there may be a need to also communicate resource and money requirements in the form of: • • Resource plots; S (resource) curves Budget, cash flow; S (money) curve.

and also possibly delay information. End-users are acquainted with their individual responsibilities, possibly through responsibility matrices. See Chapter 3 for the descriptions of these representations and terminology. The program is the visible part of the planning process, used to communicate the results of planning. Programs are guides to assist the project work being carried out efficiently. They show the steps involved in the project work. Project participants can use the programs to see where they fit in the project. Durations and start dates might be arbitrarily imposed or decided in cooperation with all stakeholders, including contractors and equipment suppliers. Programs give referenced (as-planned) baselines against which actual progress can be compared. Programs are continually updated, throughout a project, to reflect the latest information available. Sometimes bad practice is seen whereby a program becomes an end in itself, and is not used to assist replanning and project management. For large projects, the number of activities may become unwieldy to handle together, and it may become necessary to create a number of convenient subprojects, or to group activities. It is possibly too obvious to say that programs should be appropriate to the relevant end-user, and that they should be clear, user-friendly, at the appropriate degree of detail, and uncluttered in order to communicate the agreed plan. Some planners use the K.I.S.S. (keep it simple stupid) principle in order to improve understanding. Complicated programs are not recommended. The maintenance of the program may be the responsibility of the contractor. Subcontractors may have a say in where they fit in the program, or may be advised when they are required to appear. Definiteness on these appearance dates assists subcontractors to manage their overall affairs. Vagueness can lead to a falling out between contractor and subcontractors as each may have a different view on what is expected of the subcontractors.

140 Conventional treatment from a systems perspective

Programs over levels On major projects, the communication of planning outcomes in forms such as programs could be expected to occur on several levels. The levels roughly align with organisational levels. Senior management may only be interested in broad issues. Middle management would show interest in more detail. Individual production groups would work with still more detailed information. There is no consensus as to the number of levels used or what each level is called. Figure 4.9 shows three levels, but less or more than this may be adopted depending on the situation. The higher the level, the lower the degree of detail. With reference to Figure 4.9: • • • Upper level For senior management and the owner, a program is developed showing key work parcels and sequencing, milestones, and possibly contingencies (Figure 4.10). Middle level Middle management requires an indication of major work parcels and major events. Lower level Production management or workface management works off a program suitable for individual production groups.

Owners and senior management tend to be more concerned with overall results and not in the detail. In contrast, a foreman, say, needs detailed information over short periods to guide the workface employees. Intermediate detail may be required for contractual purposes when tendering, and may show major items and completion times.

Key work parcels, milestones

Upper level

Senior management, owner

Major work parcels, major events

Middle level

Middle management (Engineering, construction, commissioning, . . .)

Specific work activities, associated events

Lower level

Production management, workface management

Figure 4.9 Example hierarchy of programs.

The better network analysis computer packages cater for different levels of reporting. The lower levels are equivalent to subprojects. example project involving a conference. .10 Example upper level program (simplified).12 shows four program types and a range of common time horizons. Figure 4. for any significant project.) The programs at the higher levels are aggregations or integrations of programs at lower levels. cost and duration information at the lower levels is folded up to higher levels (Figure 4.The planning process 141 This year Subproject Responsibility J 1 Venue 2 Printing and Publications 3 Marketing and Advertising 4 Audio Visual Equipment 5 Registration 6 Speakers/VIPs Venue Committee/ Secretariat Secretariat F ∆ ∆ ∆ ∆ M A M J J A S O N D Next year J F M Venue ∆ milestone Secretariat Committee/ Secretariat Secretariat Secretariat 7 Accounting 8 Sponsorship and Exhibition 9 Social Program /Meals ∆ ∆ Secretariat Figure 4. Program time horizons For communicating the outcomes of planning.12 is based on the assumption that early in the project and throughout the project there will be a need for an overall project program (longest time horizon program) showing major work parcels. sub-subprojects. Resource. (Work parcel names have been abbreviated because of space limitations – work items are implied. milestones and key resources. Figure 4. there is commonly a need for a number of programs covering different time horizons and associated different degrees of detail. There is no consensus as to the number of program types or the time horizon selected for each. but different representations may be adopted depending on the situation.11).

construction of a multistorey building. trades 0 Daily program 0 Immediate work effort Time Figure 4. 1000 activities etc. beams 1st fl. subcontractors. gr–1st fl.Upper level Property acquisition Design Approvals Construction Commissioning/handover Intermediate level Excavation Footings Columns. Slab. etc.11 Different levels of programs. (Work parcel/activity names have been abbreviated because of space limitations – work items are implied. . milestones. 100–200 activities Lower level Falsework Formwork Reinforcement Concrete Curing etc. example. work method changes.) Overall project program 0 Major work parcels. key resources Medium range program 0 Tracks delays. 10–20 broad activities Figure 4.12 Example programs with different time horizons. scope changes Short range program Detail for crews.

Where the project is being fast-tracked.The planning process 143 As the project progresses there will be a need for a medium-range program showing detail over the next. and to ensure that the project proceeds speedily. 30–90 days of the project while other programming is being developed. monthly. The above time horizons are examples only. Table 4.2 attempts to summarise the various types of programs used on projects. 2–3 months and updated. a starter program may be issued along with the preliminary project program. when they are to be produced during project execution. At the workface level for individual crews. the time horizons are selected to suit. for example.2 Example summary program forms Level Upper Coverage Outline Outline/broad Extent or scale Entire project Project design. or to the number of rows in Table 4. For any given project.2. work method changes and scope changes. Execution/ implementation Execution/ implementation period 5–10 weeks 1–2 weeks 1–2 days Unit Month Week Degree of detail End-user Low Low/medium Management Middle Execution/ implementation Short term Weekly Daily Week Medium Project office Lower Day Medium/high Half-day High Hour High Workface . say. Detailed execution (fabrication. subcontractors and tradespeople there will be a need for a short range program showing detail over the next. delays. There is no consensus as to the usage of terminology in Table 4. say. Summary Table 4. what they must contain. 2–3 weeks and updated. The levels correspond roughly with the levels in an organisational structure. say. weekly. describing all programs. Some organisations go to the trouble of creating a summary document. Updates will take account of. and to whom they will be directed. This program need be no more than a list of items. construction ) programs are developed during the design phase prior to the start of any physical work. say. There is also a place for a daily program (shortest time horizon program) showing immediate work effort.2. say. The starter program lists the important activities to be performed during the first.

liaising with production as to realistic durations. The introduction of PC-based scheduling packages bought about changes (more detail) in the presentation of master schedules. monitored by both parties at formal review meetings with actions required if the project falls behind schedule. The company’s planning practices included liaising with the customer in establishing milestones. with a total duration. and total duration. . and monitoring actual progress against as-planned. the above inclusions were repeated for every product despatched over a two. was put together before the first master schedule was completed. Possible inclusions within the schedule included: • • • • • • Despatch date Paint start and finish Final assembly Carbon fibre component start and finish Metal component manufacture – in batches Raw material requirements. were reviewed continually to reflect the true manufacturing situation. A broad network diagram. During the implementation phase. The activity durations. For example.to three-year period. not 200 bars.144 Conventional treatment from a systems perspective Case example – Aerospace industry This case example discusses the use of schedules and network diagrams used within an aerospace company for the management and reporting of progress on its projects. hand drawn. schedules were shown as both broad hand-drawn bar charts and cumulative production plots (linear schedules). or milestones. The packages have enabled more detail to be included in the planning. the (broad) network diagram and (broad) master schedule were changed to reflect the new situation (ahead/behind schedule) and when the company would be back on schedule.13). The master schedules tended to only show major events. Because each completed component was expensive. and the logic of the network diagram. the manufacture of 200 metal components for use in final assembly was shown as one bar. Resource planning and cost impact were not performed adequately until the use of PC-based planning packages. Cumulative production plots were also used in broad form (Figure 4. usually shown as a milestone. The master schedules formed part of a contract with the respective customer. Formerly.

supplier. installer and operator of engineered solutions and equipment for the water and wastewater industry. Case example – Water treatment plant The company in this case example is a designer. This case example focuses on a project to develop a water treatment plant for a mine.The planning process 145 Batch manufacture of metal components and procurement of raw materials Paint and despatch Units Final assembly Carbon fibre manufacture Time Figure 4. The company has projects of various sizes and complexity and in various stages of completion underway at any one time. without the critical path being known. amongst other things. relevant people were asked to reassess the activities and when they could be started/completed. and spread geographically. The company was awarded a contract by the mine for the design. supply. Where progress fell behind schedule. This would lead. recovery schedules were derived to bring the overall project back into line. each aspect of the network diagram was reviewed. installation and commissioning of a water treatment plant to . to overlapping activities. manufacturer. Under these circumstances. Formerly. This also created a problem for the project manager. the people responsible for performing the actual work were not aware of the importance in adhering to designated schedules.13 Example cumulative production plot. Units are the quantity produced during the contract period or part thereof.

manufacture and delivery work without regard or distinction between the types. This program contained some thirty elements without background assumptions. Consider a project with which you have been involved and consider the number of levels of people involved. select. a detailed monthly progress report was to be submitted as an ongoing task. Programs are usually prepared with different degrees of detail for different end-users. as and when required. the company provided a very detailed program. while people at the workface need very detailed task-by-task information. The contract required a completion date of one year. Exercises 1. The procurement of the project valves as a total group was identified on the subprogram as engineer. the project was split and two separable portions were formed with staggered handover dates. engineering and procurement detail programs. makes and manufacturers. and identified the critical path. Under the terms of the contract. by negotiated variation. Two months after starting work. A work breakdown structure gave the fundamental elements of the work. This required a revised process and alternative equipment to provide for the increased flows.146 Conventional treatment from a systems perspective supply demineralised quality water. How many programs would be necessary and what degree of detail would go into each program? . order. For example top management requires broad overview information. Each report contained an updated program with progress reported against activities and the S curve. As the project unfolded. enquiry. A revised program detailing the new process and methodology was submitted. the mine increased the output of the plant. The subprogram catered for all the variables and delivery of the selected valves to the site. together with derivative programs such as four-week look ahead. many special in-house subprograms were produced as working tools. The program consisted of some 400 activities. this time with some 500 activities. size. there were no separable portions. Immediately upon contract award. As part of its tender submission the company provided a program in simple bar chart format setting out the fundamental elements and work order of the project. Procurement programs for items such as plant valves were developed. Again the critical path was identified and an S curve (money) given. An S curve (money) was also included. As a result of the process change.

[∗ Program end-users refer to those who are to use the plan. in your opinion. Rank. For a project with which you are familiar. that you may be aware of. cause the greatest detriment through to the least detriment. what would be the most important tasks that should be incorporated in a preliminary project program/schedule? 4. What are the causes of these practices? A cause–effect diagram might be helpful in answering this question. constraints ) Completion times given based on inadequate data. the so-called ‘doers’.The planning process 147 2. What suggestions would you make to eliminate the practices? 3. and also rank all the practices that.] Add others to this list. resources. How many programs were necessary and what time horizons were used in each program? . in your opinion. Consider a project with which you have been involved and consider the time horizons of the programs involved. the following poor practices in programs: • • • • • • • Too much or too little information Development without foresight as to the program end-users∗ Lack of continuous updating The use of unsuitable computer packages Lack of identifying programs as part of contract administration procedures Lack of support of programs with planning information (assumptions on methods.

The same holds true where multiple projects are combined on an organisation level. Subprojects may be chosen such that the interaction between subprojects is minimal and is commonly neglected.1): • The project (work) is broken down (hierarchically) to lower level subprojects. says nothing of interaction (or interrelationship) between subprojects. Such a work breakdown can be represented as a tree-type diagram called a work breakdown structure (WBS). to make the level problems less complicated. Interaction occurs because of the work method (implying an order/sequencing of work). between sub-subprojects. sub-subprojects and so on to activities. between . . Iterations (between levels) can largely be removed by the planner being aware of what is happening on other levels when solving an individual level problem. Planning goes from the general to the particular. Work breakdown The work in a project may be systematically broken down to subprojects to sub-subprojects and so on to component activities. however. one of which is over levels (Figure 5. from coarse to fine (broad to detailed) as the levels descend. The levels are uncoupled as far as possible. However. for tractability reasons. interaction at levels lower than subproject is commonly strong. Detail emerges from lower levels. permitting subprojects to be planned independently of other subprojects.Chapter 5 Planning over levels Introduction Simplifications Planning is simplified in a number of pragmatic ways. The WBS. elements and constituents.

1 Simplifications to make the planning problem more tractable. Controls chosen (actions that are taken by the planner) depend on the planner’s experience and expertise.and sub-subproject-type levels (equivalently work packages) can be inserted. Similarly. but generally planners are after a satisfactory solution rather than some theoretically optimum solution (within an iterative analysis approach to planning). activities. Controls at any level contain lower level control information. elements combine to give an activity. A work breakdown also says nothing about events.Planning over levels 149 Planning problem simplifications Conversion from synthesis to iterative analysis Hierarchical breakdown of a project Conversion from probabilistic to deterministic Figure 5. subproject. Formalism Chapter 9 develops formal project models. A systematic breakdown is preferred over an ad hoc approach because of the increased likelihood of identifying all activities.2. as well as the situation and a bit of creativity. Some choices may be better than others. inserting a ‘program’ level adds nothing new. though loose practice by many writers confuses this.2 shows two possible representations. Figure 5. In Figure 5. . Constituents feed up to elements. a program is no different in character to a project. a sub-subproject and a subproject (and work packages) are no different in character to a project. Control The following examples of controls or actions are not totally exhaustive. Hierarchical modelling Levels A model of a project for planning purposes can be built from the bottom-up using component parts. activities combine to give a project (and projects can be combined to give a ‘program’). but add nothing new to the following development.

In place of more usual finish-to-start (F/S) relationships. Some activities can be carried out sequentially.150 Conventional treatment from a systems perspective Project Activities Activity model Activity interaction Project level Activity Elements Element model Element interaction Activity level Element Constituents Constituent model Constituent interaction Element level Constituent Constituent model Constituent level (a) (b) Figure 5. different techniques. • • . different ways of conducting activities. and prefabrication rather than an in situ work method. or start-to-finish (S/F) relationships might be possible. different managerial and technical approaches to the work. different work sequences or logical ordering of activities. as well as finish-to-start (F/S) relationships with lead time periods. Using overlapping relationships between activities. start-to-start (S/S). Project level controls Controls at the project level relate to the work method perhaps involving the interaction of activities (and this implies an order/sequence of work). Run concurrent/parallel activities. Examples include: • Examining alternative methods of work.2 Building up a project model. finishto-finish (F/F). Examples include outsourcing or contracting out some work instead of doing the work in-house. using machinery instead of labour. some simultaneously. This will lead to alternative network logic.

Different ways of using resources (or money) on activities. control relates to resource usage within an activity: • • Total quantity – Use additional/less resource numbers. Extending the project duration. All activities then become non-critical. lengthen activities and split activities. though this may come with lower production and a cost penalty. This may go under the name of resource redeployment or redistribution. • • • • • Element level controls At the element level. Making use of any activity float. Distribution in time – Compress activities. and only lower level controls remain to be selected. For example. there is a collection of resources (different resources and possibly different numbers of each resource type). decreased quantity or increased quantity. for example: • Adjustment of activity resource numbers (each resource type) – distribution over the activity. Adjusting resources may lead to changed activity durations.Planning over levels 151 • • Expediting materials. the controls at the project level are fixed. Lengthening activities. Compressing the durations of activities involves shortening the activity durations. Activity level controls At the activity level. Splitting activities that can be split. Splitting implies starting an activity. . for each resource type. The simpler planning problem fixes the work method and only leaves resource usage and resource production rates as the variables to be determined. Changing the work calendars. That is. Control relates to the selection of the number (quantity) and type of each resource. In conventional planning parlance. this usually comes with an additional and a more concentrated resource requirement (and an increased cost). Note that shifting the start dates of critical activities will extend the project completion date. starting the activity after a break and so on. stopping the activity. introducing ‘just-in-time’ (JIT) or other inventory practices. this is achieved through. sometimes called uneconomical drag out. Compression of activities. Typically the start dates of noncritical activities are moved.

The systematic nature of a work breakdown is destroyed by the ad hoc introduction of events and deliverables. A project (scope) can be decomposed into subprojects and then into sub-subprojects and so on down to the activity level (but usually not finer). changing the work conditions. Work breakdown General Work breakdown is a top-down approach. For example. reward) or stick. a work breakdown entry may be labelled as ‘book’ when it should be ‘some work involving the book’. This is not to say that events and deliverables are not important for project management purposes. • . which is developed from the bottom-up. The term ‘work breakdown structure’ (WBS) is used to refer to the resulting systematic decomposition (as opposed to an unstructured activity list where there is a possibility that some activities have been overlooked). for example for people. Different numbers and types of resources lead to different production. it is not work. Resource production rates are different for different pieces of equipment. having the correct tools and specialisation. as well arbitrary or ad hoc practices are to be avoided. The process of decomposing a project might be called scope delineation (or scope definition) by some writers. Two wrong and confusing practices that can be seen with some writers are: • The inclusion of events (and end-products and ‘deliverables’) in a work breakdown. The labelling of entries in the work breakdown as non-work items. What is not evident from the work breakdown structure is the degree of interrelationship between its components. A work breakdown says nothing about events. Constituent level controls The constituent level is the basic behaviour level of a resource (person. or piece of equipment).152 Conventional treatment from a systems perspective working multiple shifts or overtime fits within this meaning. or the reliance of some on others. given in the previous sections and as discussed later. the controls are the resource production rates or equivalents of each resource type. An event is the start or end of a work item. Resource production rates. and how the work breakdown structure has been used to separate individual components. can be altered by the use of a carrot (incentive. but rather events and deliverables have no place in a work breakdown. as opposed to the hierarchical modelling. Resource information can be collectively measured in a money unit. As such. The lower levels represent finer detail.

Rather. there is no set standard for numbering schemes.4. and reduces the likelihood that activities will go unthought of (thereby avoiding disruption to the project rhythm.4) that recognise the hierarchy involved. Representation Work breakdown may be represented as a tree diagram (Figure 5. with money. the incurring of extra duration and resources and hence extra cost. . or numbering for accounting purposes. It enables a clear understanding of the scope. for example.3 Work breakdown structure (tree diagram form). With respect to Figure 5. and the consequent effects on team morale). the schemes may reflect different project characteristics and different organisational ways of dealing. or as in Figure 5.Planning over levels 153 The work breakdown structure provides a common framework for dealing with all aspects of planning a project (in an iterative analysis sense) – a divide-and-conquer approach. Example Consider a project involving the staging of a conference.5. or appropriately indented dot points. Activity Figure 5. . appropriately numbered items (Figure 5. Major subprojects would centre on work associated with: • • • • Venue Printing and publications Marketing and advertising Audio-visual equipment Project Subproject Sub-subproject . .3).

The tree . Figure 5. • • • • • Registration Speakers/guests Accounting Sponsorship and exhibition Social program/meals. 1300 etc. Figure 5.5 Hierarchical decomposition of a project. 1200 Sub-subproject 1201 Activity 1202 etc.4 Example hierarchical breakdown and numbering of activities. 2000 Subproject 3000 etc. These may be further subdivided into their component parts.6 shows how this may be organised hierarchically in a tree diagram. Project Subproject/ sub-subproject/ work package/ jobs/ subjobs Activities Figure 5.154 Conventional treatment from a systems perspective Project 1000 Subproject 1100 Sub-subproject 1101 Activity 1102 Activity 1103 etc.

All the work that has to be done to complete a project is contained within the work breakdown structure. the way the work is to be carried out Contracts Region. the collective bottom level of activities should be the same. The particular subdivisions chosen in the breakdown will reflect some convenient way of parcelling and doing the work. decoration activities Figure 5. but this is not necessary. time period phasing Work type or nature. diagram (Figure 5.6) says nothing about the order of and the interaction among the work. may choose to subdivide it differently. cost centres. Subproject choice Subprojects may be chosen at the convenience of the planner but might follow a breakdown according to: • • • • • • • • • Phase (chronological). publications work Venue work Hire activities Security activities Display. Irrespective of the breakdown adopted. activity groupings Function. location or geographical area Organisational breakdown (Existing) cost codes. . if given the same project.Planning over levels 155 Conference project Marketing. resource groupings Work parcels. It can be helpful to list the subprojects in the order in which they might normally be expected to occur on the project. project cost and data summary reports The nature of the planning network. cost headings.6 Tree diagram conference example. advertising work Printing. Different organisations and people.

Complete decomposition may not occur in some instances. resources and cost can be given at each level. . others may be defined only to a level where a general understanding of the activities is evident. What is ‘adequate’ for estimating purposes may change over the course of a project. At the lowest level of this hierarchical breakdown of a project the entities are manageable components. namely individual activities. For example. such that information on durations. progressive measurement and reporting relate to these activities and estimates. costing. while each trade contractor will subdivide its work package to smaller items. This lack of complete breakdown may lead to some problems later in the project with some planning and design information being found to be incomplete. comes an aggregation of duration. Lower levels can be integrated or aggregated (folded up) to higher levels for reporting or communication purposes. resource and cost information. uncertainties. Until the breakdown reaches the level of activities. resource and cost estimates. reporting (planned and actual production/progress and resource usage comparison) and replanning can be carried out at each level. responsibility and authority can be assigned to each. as the project becomes more definite. Number of levels The number of levels needed to describe all the project work might range from about two to about five. a project manager may stop at the level of trade contract packages. the activities are able to be characterised adequately in terms of duration. scheduling. Resourcing. This may be due to the complicated nature of these components or further breakdown may be considered unnecessary.156 Conventional treatment from a systems perspective Commercially available computer packages used to assist planning commonly adopt a work breakdown structure as a basis for data input. Different stakeholders may choose to decompose to different levels. the development of a work breakdown structure may proceed simultaneously from the top-down and from the bottom-up. Monitoring (project production/progress and resource usage). The lowest level is chosen to give manageable size entities or identifiable pieces of work. In practice. it is difficult to undertake estimating accurately. While certain areas may be well defined and broken down into clear and measurable components. These are of relatively short time span. The breakdown may evolve as the project progresses. Any level is the sum of the immediately lower level work. At intermediate levels the entities are of longer time span. With this aggregation.

Einstein is said to have remarked that: Everything should be made as simple as possible. The second category represents scope changes.7) – project team structure when applied to a project. and contributions by different authors. or organisational breakdown structure (OBS) when applied to an organisation and its functional lines – a representation of the personnel involved and their interrelationships. Example – Writing a book Consider the writing of a book involving the researching of numerous sources. Change Change may be viewed as: • • Changes to the nature of activities within the work breakdown structure and Required additional/less activities. though from a contractual viewpoint might be regarded as a variation. A breakdown of the work .7 Possible relationship between work breakdown and organisational breakdown. but not simpler. Changes to the nature of activities within the work breakdown structure may not normally be regarded as scope changes.Planning over levels 157 Project manager Project Organisation Project team Subprojects Divisional responsibility Work packages Departmental responsibility Activities Individual responsibility Figure 5. WBS relationship to OBS The project breakdown may bear a close relationship with organisational breakdown (Figure 5.

9. Figure 5. Example – Building a house Consider the building of a low-cost house. sentences. resource and cost estimates for the writing of each chapter/section/subsection can be made. section x. The development of the WBS proceeds from the top-down but relies on the end-users (writers) providing input of what is required at the lower levels. might be viewed according to Figure 5.xx Writing etc.158 Conventional treatment from a systems perspective Writing etc. Duration. The component activities might be grouped as in Figure 5. Cost account numbers could parallel the chapter/section/subsection numbering scheme.10 shows schematically this breakdown. Example – Road project A breakdown for a road project might look something like Figure 5.8 Example WBS for compiling a book. resource and cost estimate for the book.00 Writing etc. Monitoring and replanning the writing of the book can be directly related to chapters/sections/subsections. words and letters could be considered but are perhaps too small to be manageable. .11. sentences. A responsibility matrix could be set up identifying each chapter/section/subsection with a writer.x0 Writing etc. Lower subdivisions to writing paragraphs.8. checker and editor. words. subsection etc. and a schedule and budget for its writing. book x. letters Figure 5. chapter x. writing paragraphs. aggregated to give an overall duration.

1 1.1 Laying driveway. install septic tank 6. .2 Placing roof tiles 5.1 5.1 Erection of site facilities 1.2 5.1 Setting formwork and reinforcing steel 3.1 2 3 4 5 6 7 Preparatory/preliminary work 1.2 Installation of door and window frames 4.2 4 4.1 3.4 Running electric wiring Other work 7.1 4.9 Example work breakdown for the building of a low-cost house.3 3 3.3 Plastering 4.2 Concreting Wall construction 4.2 Gardening 7. footpath 7.2 4.2 2 2.3 6 6.1 2.3 4. House building 1 1.10 Example work breakdown for the building of a low-cost house.3 Figure 5. sewerage pipes 6.1 Excavation 2.3 Laying water pipes 6.4 Painting Roof/ceiling construction 5.1 Excavation.1 Roof framing 5.2 7.3 Concreting Floor construction 3.1 6.2 Survey and layout site Foundation work 2.2 2.3 Installing ceiling Plumbing/electrical work 6.3 Cleaning up Figure 5.2 Laying drainage.1 Bricklaying 4.4 5 5.1 7.2 Setting formwork and reinforcing steel 2.2 6.4 7 7.3 6.

from the viewpoint of the owner of the mine and the trucks. in the breakdown that follows. blast Transport Fill Base. all constituted initial stages of planning.160 Conventional treatment from a systems perspective Road project Survey work Earthworks Excavation Drill. Issues relating to transmission life 3. transmission and differential life.11 WBS for an example road project. There were many issues which affected the costs of haul trucks. A problem was identified and a project developed to investigate that problem. A research and development project was established. The mine was uncertain as to why these trends were occurring and hence was required to do further research. and in particular relating to brakes. After preliminary research into the trucks. Procedures used and decisions made. it was noticed that the majority of costs incurred were in maintenance. The overall aim from the owner’s view was to reduce costs associated with the operation of the haul trucks. The case example relates to a research and development project established due to inadequate brake. it was first divided into research work associated with three subprojects: 1. The mine looking after these trucks had to take the research results and address them wherever practical. honing in on . bridges work Figure 5. Issues relating to brake life 2. transmission and differential life of mining haul trucks. Case example – Truck research This case example describes the work breakdown structure for a project involving the examination of the operation of mining haul trucks. Issues relating to differential life. Research work is implied. but not stated. seal work Culverts. up to and including this (first) division. Due to the expansive nature of the project.

then they needed to track back to this first division and follow a different path.Planning over levels 161 maintenance issues already created a subproject. through a systematic approach backed up by previously collated data. honing in on the correct components that were of the greatest cost would not have been possible. If work breakdowns had not occurred. because there were other ways that costs on the haul trucks could be addressed. but rather what was costing. it was still too cumbersome to allow for factual and clear communication of identified problem areas.1 Factors relating to the application of the trucks 1.1. Within each of these. There were many factors which may lead to limiting the brake life of these trucks at this level. a further breakdown was evident. A sub-aim of this project was to identify the limiting factors of brake life for the haul trucks. experience etc. this was further broken down into three distinct areas based on truck components. then identifying an area where most expense on the trucks was occurring would have been a burdensome task. if this path was proved to be wrong. At this point. limiting factors and therefore areas with associated cost were to be identified and were to help contribute to answering the overall aim. The ‘why’ details came forth later in the project breakdown.1 1. Two clearly separate issues relating to research on brake life were identified (second division): 1. The mine had effectively narrowed down an area based on collected data.2 1. that is to sub-sub-subprojects.1. It also followed.1. so it was once again broken down smaller.3 Speed in relation to brake effectiveness Haul road profiles and effect on braking system Loading of vehicles . that had the first division not been backed up by history files. where they thought that significant cost savings could be produced. creating subprojects helped to keep the overall project’s aims in sight whilst breaking the project into subprojects and sub-subprojects enabled easier management and planning. a new set of aims (sub-aims) was defined. this next breakdown addressed research work areas that were categorised according to the above factors (third division): 1. However.2 Factors relating to the design of the trucks.. Not much effort was spent looking into the detail of why or how the problems were occurring. Narrowing in on the issues relating to brake life. to assist in keeping the overall project’s aims on track. Once this sub-aim was satisfied. Even though the subproject had been divided into sub-subprojects to have reached the above stage. Once the overall project had been given a definite subproject to follow.

4 1.6 Operator considerations/human issues Maintenance considerations.3 1. For the incrementally launched box-girder bridge.1. General agreement and past experience between the planning .2. This division was based on brainstorming all possible contributing factors to brake life.2. the breakdown chosen stemmed from a combination of planning requirements and the ease by which the project manager could monitor. The second division was made by the subproject manager so that the sub-aims could be clearly met. The work procedures highlighted how the third division aims contributed to the overall project’s aims. The procedure followed to come to the second and third divisions differed from that followed to come to the first.162 Conventional treatment from a systems perspective 1. Characteristics of fluid in hydraulic system Material of brake pads Design characteristic monitoring Different types of braking systems Environmental effects Pressure system and fluid flow for hydraulic brakes.2. The second division and the ones that followed were based on hypothetical possibilities.1 1. Future steps then became specific activities rather than areas and so on.1. After brainstorming it became evident that two main factors could be contributing to the brake costs. These were seen as one of the last steps in the breakdown tree. manage and record construction activities and costs associated with each subproject.2 1. Case example – Bridge construction This case example describes the work breakdown structure (WBS) for the construction of a bridge from the viewpoint of the contractor. This was because these later two divisions were the initial focus on not only what could be contributing to the brake life but also why these categories could be contributing to the brake life problems.4 1.5 and 1.2.5 1. The items which fell into these two main categories went together to make the third division.2. there was no firm previous data that could have been collected to give the best direction.2. Activities and work procedures were then identified and were combined in addressing the set of aims associated with the third division.

The breakdown of the project into these four major subprojects was intended both for ease of record keeping and for usefulness of the records kept. information on the costs of building that particular type of box girder. is useful and hence was recorded. by separating the substructure work. Whilst box-girder bridges may have different piling (substructure) and pier column arrangements. The work breakdown structure also followed the course of the job. safety. records were now available on the costs and production of the barge. consumables. the piling hammer and the manufacture of pre-stressed concrete piles. regardless of the foundations. Separate work breakdown structures existed for activities associated with site mobilisation and demobilisation. Similarly. electricity. The work breakdown structure for the project consisted of four main subprojects. Schedule management was achieved by being able to systematically close-out subprojects.Planning over levels 163 and construction teams limited the number of work breakdown levels. site overheads and so on. Each of these subprojects was further divided into work associated with: • Plant and equipment – Hire – Maintenance • • Concrete Reinforcement – Subcontract – Materials • Formwork – Subcontract – Materials. tools. This enabled bridge records to be used for a variety of other projects. for example wharves. in both schedule and cost senses. namely work associated with: 1 2 3 4 Substructure Piers and abutments Box girder Completion of superstructure. the project manager was able to monitor the project. By relating the work breakdown structure to the master plan. Cost savings in a completed .

the stakeholders understood their roles and those of others within the project. materials and labour was largely a breakdown based on costs. the greatest variable. thus the breakdown assisted in the preparation of monthly accounts. The majority of plant and equipment for the project was obtained on internal hire. it was assumed by all stakeholders that ‘all bases had been covered’. The project. The first three of these were achieved with only relatively minor troubles and concerns. Everyone. the work breakdown structure (WBS) was not clearly understood by the stakeholders. and subsequent blow-outs in this area pointed to estimating problems. Labour costs. The subsequent breakdown into work associated with plant and equipment. resulting in confusion. Other companies were working on the design of the document management system profiles for the new documents to be created and their conversion. At the completion of planning. Various companies were working together on the document macro and template developments and design of a conversion filter for existing documents. frustration and near disaster. Materials were assumed to be estimated accurately at tender stage. was actively engaged in working towards converting existing documents . while the fourth proved to be a major headache due to the number of subprojects it contained and the necessity to outsource five of these subprojects to five separate organisations – a decision based on a lack of required expertise and experience within the firm. it seemed. gave an indication of the management of the project. and nothing had been overlooked.164 Conventional treatment from a systems perspective subproject could be used to offset the increased costs in another subproject. and by the time it came to update the application software. involved: • • • • • Upgrading all hardware Upgrading all hubs and associated cabling Upgrading all operating systems Updating application software Training of all staff in the new software applications. with its end result of upgraded technology throughout a firm. Case example – Computer system upgrade In this case example project. enabling the project to remain within overall budget.

no one had been assigned responsibility for the actual conversion of the existing word-processing documents. offsite. therefore. without being able to really manage . However. several other activities were to occur concurrently: the rolling out of the new desktops onto their PCs. the person who was expected to be the main impetus behind the success of the project was so busy and so much in demand that. who nevertheless expected him to add it to his normal duties. It would have been impossible to convert their documents across in time for them to work for the first few days after training. Their document profiles would have been converted to the new document management system and their PCs would have been ready with the new system but they would not have been able to find any existing work. was constantly spreading himself around in smaller. the conversion of their word-processing documents and their document management profiles. this was a situation which would have had serious repercussions. The project had continual changes within each subproject. with good people and communication skills. this error was discovered the day before training was to commence. Had their documents not been converted upon their return from training.Planning over levels 165 and document management system profiles over to their new format. A few faxes confirming verbal quotations were all that were present. Each stakeholder assumed that another was responsible for this activity because everyone had seen and worked with the conversion filter at some stage and had discussed it at some point during the project. He also had the support of senior management. soon enough to avert a near disaster. As a result of this. superficial time lots. Training was to be in groups of six staff members at a time. they would not have been able to access critical work. The project manager was appointed from within the firm. instead of being able to dedicate a worthwhile period of time and attention to ensuring that each activity was performed. he had the necessary skills to understand the scope (of the project). should it have remained undetected. and preparing for the creation of new documents and profiles. While each group was in training. It was assumed by all. and because thousands of these documents were vital legal documents. that preparations for the commencement of training were in hand. Subproject work that was done by outside contract lacked a scope statement defining the extent of the work. but even these did nothing to state the overall area of responsibility or scope (of work) to be undertaken by each contractor or the firm. and was a good manager. in all aspects of the new system. It could have been easily avoided by better planning and closer attention to detail. as the software was developed. Fortunately. Despite having no prior project management experience.

thereby taking them away from other work. Iterative analysis approach Generally. possibly.12).13. It may have drawn attention to the impending problem. cumulative production plots and so on. accusations and.166 Conventional treatment from a systems perspective the project. The analysis and evaluation will generally involve network analysis and the use of time-scaled networks. scaled models and simulation. experience and knowledge of the industry.9. arguments. showing the feedback loops. bar charts. . and in particular the feedback loops between levels. Perhaps if such progress had been documented. As a result of this limit on human resource numbers. with activities delegated to appropriate members of the team. The iterative analysis approach follows Figure 5. Figure 5. have included a detailed WBS. with the reason that ‘no more resources (people) could be spared’. Had the documents not been converted in time. One of these was that documentation of the progress of the project was sparse.13 is saying the same thing as Figure 2. It also would have caused a blow-out in the schedule and budget because the firm possibly would have had to then insist that some of the trainees get involved in the conversion process. Planning problem simplifications Conversion from synthesis to iterative analysis Hierarchical breakdown of a project Conversion from probabilistic to deterministic Figure 5. When he requested assistance from senior management it was refused. it would then have drawn attention to the oversight on the activity of converting the documents. at a minimum. the inevitable inability to access work would have occurred for the trainees. This may have had the effect of causing frustration. The initial solution guess is commonly based on past projects. The analysis may be assisted by 2D and 3D computer visualisation.12 Simplifications to make the planning problem more tractable. primarily because the synthesis version is too difficult (Figure 5. other problems arose. planning over levels is practised in conjunction with an iterative analysis mode of attack. Documentation could.

Knowledge of the project and the industry will help develop . sequencing) Adjustment of project level controls Activity level Activity resource selection Adjustment of activity level controls Element level Element resource number. Approaches are only limited by the planner’s imagination. Controls selected depend on the planner’s experience and expertise.Planning over levels 167 Owner's values objectives. constraints. scope. as well as the situation and some creativity. constraints Adjustment of value system Project level Work method selection (inc. If one approach doesn’t work or doesn’t work completely. then other approaches may be tried. evaluation Objectives. work breakdown. Generally planners are after a satisfactory solution rather than some theoretically optimum solution. performance – satisfactory Objectives. constraints.13 Iterative analysis approach to planning over levels. type selection Adjustment of element level controls Constituent level Constituent production rate selection Adjustment of constituent level controls Analysis. performance – not satisfactory Planning complete Figure 5. Some solutions may be better than others.

The simpler iterative analysis approach (compared to a synthesis approach) is presented. Example To assist an understanding of planning over levels. but these could be expected to reflect the owner’s values as well. Suggested constraints might be: • • • • Availability of expertise and people Duration upper limit Cash flow. The project is the means by which this report comes about. Objectives and constraints The project objectives and constraints are from the viewpoint of the consultant undertaking the project (undertaking the work for an owner). The procedure in planning the project over levels is developed. an example is presented. amongst other things. technical or logic constraints affecting work. Whether any of these approaches are possible will be determined by the project and the nature of the project work. End-product and project An investigation report is required to be produced by a consultant. perhaps requiring payment from the owner as soon as possible Fitting in with other projects. including the usage of personnel. Any control adjustment.168 Conventional treatment from a systems perspective alternative approaches. for example. Suggested objectives might be: • • Minimum cost Maximum efficiency in work practices. may: • • • Lead to new cumulative resource (money) plots Alter the critical paths Lead to rescheduling of activities. The example project chosen is one which most people should be able to visualise. . lead to the removal of physical. Value system adjustment may redefine the planning problem or.

Planning over levels 169 Report Proposal Initiate Acceptance Literature Preliminary search Supplementary search Analysis Field data Preliminary collection Reduction Supplementary collection Reduction Analysis Hard copy Progress report Draft Final copy Figure 5. . Scope delineation – Work breakdown A possible breakdown to activities. Project level Activity interaction/sequencing Precedence relationships establish the logic of how the project is to be put together.14. is given in Figure 5. literature searches. The following is one suggested way in which the work might be carried out. (Activity names have been abbreviated because of space limitations – work items are implied. not in any particular order.14 Possible work breakdown for the investigation report example. The project proposes to use a number of people and their associated skills. It involves developing a proposal. the collection of field data and the production of hard copy.) Scope Broad scope The broad scope (of the project) is one of producing an investigation report.

Note that the network diagram is no more than the logic of how the work is to be carried out. Figure 5. (a) Activity-on-node diagram. (b) Activity-on-link diagram.15 Example project networks. Prered Suplit.15a is an activity-on-node diagram. Predat Predat Prorep Prorep Supdat.15(a or b). Figure 5. The network may be either an activity-on-link diagram or an activity-on-node diagram. Prelit Prop Accpt Predat Prered Prorep Suplit Drrep Supdat Supred Anal Firep (a) Prop Accpt Predat Prelit Prorep Suplit Anal Drrep Firep d Supdat Supred d Prered Dummy activity (b) Figure 5. The network looks like Figure 5.15b is an equivalent activity-on-link diagram. Supred Anal Drrep Network formation The network for this project can now be drawn. .170 Conventional treatment from a systems perspective Code Prop Accpt Prelit Predat Prorep Prered Suplit Supdat Supred Anal Drrep Firep Activity Initiation of proposal Acceptance of proposal Preliminary literature search Preliminary data collection Progress report Preliminary data reduction Supplementary literature search Supplementary data collection Supplementary data reduction Analysis of literature and data Prepare draft report Prepare final copy of report Precedence – Prop Accpt Accpt Prelit.

Planning over levels 171 Activity level Activity duration estimates Activity duration estimates imply that some assumptions have been made about resourcing and resource production rates. A possible set of duration estimates for the activities is as follows: Code Prop Accpt Prelit Suplit Anal Predat Prered Supdat Supred Prorep Drrep Firep Activity Initiation of proposal Acceptance of proposal Preliminary literature search Supplementary literature search Analysis of literature and data Preliminary data collection Preliminary data reduction Supplementary data collection Supplementary data reduction Progress report Prepare draft report Prepare final copy of report Duration (days) 10 2 21 7 2+4 = 6 25 6 10 3 5 12 5 These values obviously depend on the skills of the estimator and the familiarity this person has with the work involved in the project. resourcing and resource production rates are related. Duration. Resource estimates The resource requirements of ‘professionals’ is estimated to be as follows: Code Prop Accpt Prelit Suplit Anal Predat Prered Supdat Supred Prorep Drrep Firep Activity Initiation of proposal Acceptance of proposal Preliminary literature search Supplementary literature search Analysis of literature and data Preliminary data collection Preliminary data reduction Supplementary data collection Supplementary data reduction Progress report Prepare draft report Prepare final copy of report Professionals/day 2 4 2 2 1+1 = 2 3 2 3 2 1 3 1 .

Time-scaled network Figure 5. Bar chart The information from a network analysis is readily transferable into bar chart and time-scaled network form (Figures 5. Time-dependent costs have not been included. Element and constituent levels The controls at the element and constituent levels are assumed fixed for this example.17 shows the bar chart plotted for all activities starting at their earliest start times.18). and a further $120 000 be paid on completion of the final report (Firep). Figure 5. .18 gives a time-scaled network.172 Conventional treatment from a systems perspective Money estimates The following are estimates for costs to undertake the activities: Code Prop Accpt Prelit Suplit Anal Predat Prered Supdat Supred Prorep Drrep Firep Activity Initiation of proposal Acceptance of proposal Preliminary literature search Supplementary literature search Analysis of literature and data Preliminary data collection Preliminary data reduction Supplementary data collection Supplementary data reduction Progress report Prepare draft report Prepare final copy of report Cost/day ($/day) 1500 3000 1200 1200 1200 3000 1200 3000 1200 600 1800 750 The agreement for payment (income) for the work is that $120 000 be paid after the completion of the progress report (Prorep).16.17 and 5. Analysis Network analysis See Figure 5.

rostered days off (RDO). Calendar For use by project personnel. number of shifts per day and so on.16 Example network calculations.Planning over levels 173 Key : 12 33 16 37 4 Prelit 21 42 49 48 55 6 Suplit 7 42 52 42 52 0 Supred 3 52 55 52 55 0 EST EFT LST LFT TF 0 10 0 10 0 Prop 10 10 12 10 12 0 Accpt 2 37 42 37 42 0 Prorep 5 61 73 61 73 0 Drrep 12 73 78 73 78 0 Firep 5 Predat 25 12 37 12 37 0 Duration: days Prered 6 37 43 46 52 9 Supdat 10 Anal 6 55 61 55 61 0 Figure 5. 0 Prop Accpt Prelit Predat Prered Prorep Suplit Supdat Supred Anal Drrep Firep Scheduled duration Float Connection 10 20 30 40 50 60 Days 80 Figure 5. the information is more desirably presented on a calendar allowing for stipulated work hours per day. weekends.17 Connected bar chart for example project. . public holidays.

and a cumulative expenditure plot or S curve – over the duration of the project are given in Figure 5.19 shows the resource plot for the resource of ‘professionals’. With income superimposed.21 results. 6 Professionals 4 2 0 10 20 30 40 50 60 Days 80 Figure 5.20.18 Example project time-scaled network. .174 Conventional treatment from a systems perspective 0 10 20 30 40 50 60 Days 80 Suplit Anal Drrep O O Firep O O Prelit Prorep O O O Supred Supdat O Prered O Prop OO Accpt Predat Scheduled duration Float or dummy activity Figure 5. Resources Figure 5. Money Plots of money requirements (costs) – expenditure plot. Figure 5.19 Example resource plot for project ‘professionals’.

.6 Expenditure $000 4 2 0 100% Cumulative expenditure 0% 0 10 20 30 40 50 60 Days 80 Figure 5. 100% Cumulative expenditure Cumulative income 0% 0 10 20 30 40 50 60 Days 80 Figure 5.21 Example cumulative expenditure and cumulative income plots.20 Example expenditure (upper diagram) and cumulative expenditure (lower diagram) plots.

then something has to be done to reduce the peak resource requirement.. No two tree diagrams will be the same. should the availability of professionals be fixed at five people.176 Conventional treatment from a systems perspective Evaluation against objectives. It depends on the machine or toy and the way it is put together. Experienced planners may only need one. room preparation. The bottom level represents the individual nuts and bolts and discrete parts. Assume that you are undertaking a project of holding a technical or public seminar. payment. Conventionally resource smoothing addresses this problem. whiteboard ) . but is no more than an adjustment of the controls. Exercises 1. constraints and performance The results of the analysis are now evaluated against the project objectives and constraints. Repeat The whole process is repeated until the objectives and constraint satisfaction are to the liking of the planner. booking. payment ) Venue (selection – size. Some of the more important activities would involve those related to: • • Attendees (informing. the resource usage as depicted in the resource plot may not be considered efficient. The objectives and constraints may also be modified if the analysis indicates something unusual. refreshments and meals. etc. For example. handling queries. There may also be feedback to adjust the project owner’s value system if the current objectives and constraints are unsuitable in some way. Inexperienced planners may need many iterations. As an another example. Conventionally this is called resource constrained scheduling. resources and money? Feedback – Adjustment of controls The controls are adjusted at the relevant levels based on this evaluation. enrolling. How is the project in terms of schedule/production. 2. At the top of this diagram is the toy or machine. or even zero. Draw a tree diagram for the assembly of a machine or toy with which you are familiar. The next level represents subsystems. but again it is no more than an adjustment of the controls. iterations. the next again level sub-subsystems and so on. audio-visual facilities.

Draw an organisational chart identifying all the people involved. What is the difference between an unstructured activity list and a WBS? Why is the latter preferred? 5. Draw a responsibility chart/matrix (Chapters 3 and 7) concerning all the project participants. How do you know what is the lowest level to decompose a project? 4. These might include: • • • • • Speakers Company director Project manager Assistants to the project manager Attendees. Develop a work breakdown structure for this project. distribution ). the way a budget might be developed for this project.Planning over levels 177 • • • Speakers (selection. fees ) Proceedings/seminar notes (writing. A work breakdown structure (WBS) is chosen to suit the needs of each project. money and duration that are necessary to accomplish the activity. Consider now. accommodation. it is possible to not be systematic and describe the work in an unstructured list. resource and cost estimates of the project to the work breakdown structure. Develop a work breakdown structure for the project. A work breakdown structure subdivides a project into hierarchical units of subprojects work packages activities. design. However. . Relate the organisational structure that you might use to the work breakdown structure that you have just developed. certificates ) Promotion (market establishment. distribution. binding. 3. All of this has to be done under various schedule/production. It is an accepted way of representing how a project is broken into finer components. Each activity is of a size to facilitate estimating the number and type of resources. printing. resource and cost constraints. The end result of the work breakdown process is a collection of work units or activities of relatively short duration and of manageable size. 6. Relate duration. Relate the work breakdown structure to the responsibility chart/matrix and organisational chart. Consider a project associated with a social or sporting group to which you belong. transport. advertising. How could the budget format relate to the work breakdown structure? How do you make duration estimates to carry out each of the earlier mentioned activities? Consider now the personnel involved.

Consider a project associated with a social or sporting group to which you belong. money and durations can be made? .178 Conventional treatment from a systems perspective Each activity is a meaningful job for which individual responsibility can be assigned. What is the lowest level at which estimates of resources.

2004). Information assisting the replanning problem at any stage is better less far into the future. decisions could be expected to have the greatest effect. Replanning updates the controls based on actual performance. Planning always proceeds from the general to the detailed. During project execution.Chapter 6 Replanning Introduction Replanning is carried out according to project stages or at discrete points in time. Planning carried out early in a project accordingly could be expected to provide the best opportunity to determine the direction. In Figure 6. As well. A good plan allows for some flexibility and can be adapted to changing circumstances that may occur during a project. This implies continual replanning. The ability to influence the outcome of a project could be expected to diminish as the project progresses (something like Figure 6. As the project progresses. Replanning is no different in form to planning. However. it is considered important to get it right at the start. cost and so on for the project. more information becomes available. The (re)planning problem (for the remainder of the project) becomes better defined as the project progresses. the current project state and some forecast of future influences. Conversely. attention centres on keeping up to date with the effects of changes. Early. ‘unknowns’ become ‘knowns’. but with updated information. changes made later in a project are viewed as more costly. and the (re)planning can become more detailed and more accurate. and controls (immediate or contingency) developed for them much like in risk management (Carmichael. the opportunities disappear. as the project progresses. trends only are shown and the location of . delays and unforeseen events which on many projects appear inevitable. To attempt to plan in great detail right from the beginning ignores the reality that future events are difficult to predict with certainty and outcomes never eventuate as intended.1).1. The level problems repeat for each stage. Planning starts with broad assumptions which are refined as the (re)planning progresses. the planner’s experience and understanding of the project improves. As the project takes shape. attempts are made to foresee potential influences in upcoming stages.

Disturbance takes the project behaviour away from the current baseline. and uses the current project state as initial conditions. Monitoring involves measuring or observing the output or performance of a project. 2004). reporting is made appropriate to the level of management concerned.2). How data could be obtained to substantiate the widely held beliefs of Figure 6. As one precursor to replanning.180 Conventional treatment from a systems perspective Concept High Development Implementation Termination High Ability to influence cost Ability to influence cost or Savings potential Cost of change Cost of change Low Project start Time Project end Low Figure 6. Replanning incorporates updated estimates of future influences. Variances are used to guide the choice of which controls need to be adjusted to correct the path of the project. Planning. Monitoring and reporting follow. monitoring. and the cycle continues throughout the project. Disturbance takes the project behaviour away from the current baseline. Reporting is in absolute terms and terms relative to planned baseline or target performance (variances – the differences between planned and actual performance). .1 is unclear (Carmichael. Replanning gives an updated project baseline or target performance. Information desirably is presented in a simplified and easily understood form. state) of a project is monitored and the information obtained converted into reports on which management decisions (selection of controls) can be made (Figure 6. the current output (performance. 2004). Reporting specifics can take various forms. reporting and replanning Planning establishes a project ‘baseline’ or ‘target’ performance. the intersection point of the trend lines is not important. Like programs.1 Anticipated cost influence and changes over a project’s duration (Carmichael.

2 Plan–implement–monitor–report cycle. weather disruptions. all the discussion on planning applies to replanning. disciplined replanning is recommended practice. variations (design changes. resource variability (within and between activities) and so on. It is planning carried out subsequent to any original planning. Continual monitoring and feedback of project progress is necessary in a changing project and environment (reflected by disturbance). In principle. A performance measure is in the lay usage sense of the word ‘objective’ (mentioned in Carmichael.Replanning 181 Future influences estimate Controls Plan/ Replan Disturbance Implement Report Absolute and comparative performance Monitor Figure 6. replanning is no different to planning. it could be expected that the project will more likely perform close to that desired. One measure of a project’s success may be in terms of the difference between this planned and actual . 2004). this may be translated to cost). and actual performance is compared with planned performance. owner priority changes. As such. industrial disruptions. ). Planned-for project states may not be reached because of delays. else replanning may reduce to something akin to crisis management. (Project) scope also rarely stays constant. Systematic. Performance measures Performance measures are gauges by which the success or otherwise of a project is measured. All these things are collectively referred to as disturbance. Generally a project has a planned-for performance (resource usage and production over time. Generally the greater the frequency of monitoring. even though allowances for these matters may have been included in the original planning.

(But comparisons between actual performance and other performances. and unexpected wet weather arose. objectives are developed very early. The tender documents for the line construction were distributed just prior to obtaining the government approval to proceed with the project. A performance measure might be termed a ‘key performance indicator’ (KPI) or similar.182 Conventional treatment from a systems perspective performance. which meant that there were two months less available than originally estimated to construct the powerline. namely the laying of a pipeline in a main mine access road reserve. but all of these types of terms are used very loosely by most people. for example industry standards/benchmarks. The road was also the main access route for the powerline construction. materials procurement. competitors’ performance or previous projects. and a section off the mining lease area. but lack precision. and this made delivery of the . The actual construction was split into two parts. Case example – Powerline construction This case example looks at compression practices to meet a powerline construction project’s deadline. weather permitting. because of the terrain. One of the major uncertainties of the whole project was the delivery of the concrete poles to site. So the owner was able to interview the final contenders and advise them of the project’s requirements.) Performance measures tend to be developed later in projects (after the initial planning has been done). The contractor was given the responsibility for arranging and coordinating the delivery of the concrete poles with the pole supplier. design. This project started with a route selection study. The terms are used frequently by project personnel because they sound good and impress. such as long delays in obtaining government approvals. are possible. when factors external to the project. other infrastructure work for the mine was being conducted in the vicinity of the line. with emphasis being placed on the project’s deadline and their proposed work schedules to complete the work as required. environmental impact statement (EIS). Governmental approval was not obtained until a year later. Crucial to the success of the project would be close and constant monitoring of the construction activities to ensure that the project deadline was met. At the same time. in this case to the pre-drilled hole sites. survey and the letting of the construction contract. a section of the line on a mining lease area.

This meant that poles were being dumped at specified sites. This meant the contractor could utilise the additional hours.00 pm. This was due to the fact that it had not started pipe stringing as yet. Then Murphy’s Law came into play. and all stakeholders were starting to become concerned.00 am and 5. the owner started to look at what could be done to obtain some slack in the program to ensure the project would meet the required deadline.Replanning 183 poles very awkward in the majority of instances. This meant some staffing re-arrangements. but it seemed to work effectively. and this