Lecture 2a

Program Evaluation and Review

Technique

Dr Anis Rahman

G09 1.19
a.rahman@griffith.edu.au
 PERT & CPM
Scheduling

Two commonly used method for project

planning and scheduling are PERT and CPM.
Both use critical path to compute project
duration, early and late time, and slacks
(Floats)
CPM/PERT

 CPM/PERT – project scheduling technique were

developed along two parallel streams, one industrial and
the other military.
 CPM (1957)
 Discovery of M.R.Walker of E.I.Du Pont de Nemours & Co.
and J.E.Kelly of Remington Rand, circa.
 The computation was designed for the UNIVAC-I
computer.
 PERT (1958)
 Devised for the POLARIS missile program by the Program
Evaluation Branch of the Special Projects office of the U.S.
Navy, helped by the Lockheed Missile Systems division,
and the Consultant firm of Booz-Allen & Hamilton.
 CPM and PERT

CPM PERT
 CPM is a deterministic PERT is probabilistic
approach- one time estimate is approach
used for each activity
Variation of activity is only Presumed variations in the
planned instead random activity is random
 No statistical treatment of Statistical treatment for
uncertainty uncertainty.
 Includes a mathematical Strictly oriented to time
procedure to estimate the trade element of project
off between project duration
and cost.
PERT Networks
 Why do we need PERT?
 Deterministic vs. Probabilistic duration
 Allows variability in estimating duration or
 Probability of meeting the target

 PERT considers for

 Optimistic time – how long the activity would take if the
conditions were ideal
 Most likely time – most probable time if the conditions were
normal, and
 Pessimistic time – how long the activity would take if a
significant proportions of things that could go wrong, did go
wrong
 to be considered in predicting the overall time
 Start
Frequency

Optimistic

Pessimistic
Most likely

Expected
to tm te tp
3 13
5 Activity Duration
Continuous Probability Distribution
Expected time t e= [ to+ 4tm+ tp ] / 6
 Frequency

Optimistic

Pessimistic
Most likely
Expected
to t m t e tp
Statistically speaking, the distance from one end of this
distribution to the other end is 6 standards deviations
( + 3S) (p
S = t −t 6 o )
Variance can be calculated from standard deviation as V = S2
Activity to tm tp te s v
0-1 5 8 13 ? ? ?

1-2 3 5 10 ? ? ?

2-5 12 20 22 ? ? ?

Why do we need to know about V (measure of variability)?

To measure the uncertainty of our estimation.
The larger V, the less reliable te, and the higher the
likelihood that the activity will be completed much earlier or
much later than te .
Activity to tm tp t e s v
0-1 5 8 13 8.33 1.33 1.78

1-2 3 5 10 5.50 1.16 1.36 Critical

2-5 12 20 22 19 1.67 2.78 Critical
Time of the all activities in the critical path
Te = Σte
Te is computed as the sum of average times,

The variation of the distribution of Te is computed as

V (Te) = Σ V
the sum of all C.P. variances.
In other words, use the variance of critical activities to
assess the variance of the overall project.
In our analysis, we will assume that the project completion
date (Te) follows the familiar normal probability distribution
To determine what is the probability of meeting a proposed
deadline, it is necessary to determine the Z value – how many
StdDev the due date is away from the expected date.

Tx − Te

Due date (Tx)

Z=

Expected date (Te)

S ( Te )

If Tx=50 days
Te =43days
S(Te) = 5.745
Z= 1.22 standard deviation
Z
-3 -2 0 2 3
Z is the number of standard deviations the due date lies
beyond the expected date
Steps for PERT
1. Estimate to, tm, tp for each activity (given values)
2. Compute te for each activity (use te equation)
3. Calculate standard deviation for each activity
S = ( tp - t o ) / 6
4. Calculate variance for each activity
V = S2
5. Determine Te (Critical Path) = Σ te along C.P.
6. What is the confidence in calculated Te
V (te) = Σ V for critical activities only
S ( te ) = V ( te )
T −T
and project standard Zdeviation
= x e
S ( Te )
7. Find the Z value
8. Use probability tables to compute your chance in meeting the target
Te = 41.8 weeks V (Te) = Σ V = 6.86
S (Te) = 2.62
So if the project target is 43 weeks
Z = Tx - Te = 43 - 41.8
= 0.458 67.73%
S (Te) 2.62
Example 1

 Determine the probability of finishing the following project in

27 days
Activity t0 tm tp
H
A 6 14 16
2 6 I
A B 1 2 3
G
C 3 12 21
E
7 8 D 2 5 8
1
K E 6 15 30
D
F 3 4 5
J
B
G 5 7 9
3 4 5 H 4 11 12
C F
I 1 5 9
J 1 5 9
K 1 4 7
Answer

1. Step 1: Find the critical path

2. Step 2: Determine te, S, and V

3. Step 3: Estimate Te and V(te) for critical path

4. Step 4: Estimate z (area under the Normalised

curve)

5. Step 5: From z table locate the probability of

completion of the project on the due date
 Lecture 2a

Resource Planning with Network

Dr Anis Rahman

G09 1.19
a.rahman@griffith.edu.au
Resource Planning with network

 Resource constraint:
 Resources are limited and some are scarce
 Limited availability of skilled labour, equipment, even
working capital.
 This dictates that the activities should be scheduled at
time so that certain resources are balanced or
smoothed throughout the project –resource levelling
 True when multiple activities requiring the same
resources are scheduled for the same time.
 Important for multi-project organisation
 RESOURCE HANDLING
(4) (2)
0 4 4 0 8 8

FF TF FF TF
A B
0 4 0 8

6
Resource Units

A (4)
Uniform Distribution

2
B(2) Time
0 4 8
 (4)
5 8 13

FF TF
C
5 13

(2) (3) (3)

0 5 5 5 7 12 13 7 20
FF TF FF TF FF TF
A B E
0 5 13 13 20
6
Resource
(4) Histogram
ss 1 6 4 10

FF TF
D
9 13
ES LS
11 11
9 D D
7 7
B B
4 4
C C
2 2
A
A E A E

0 5 13 20 0 5 13 20
Exercise (Constrained resources)

For the following project you have only 15 workers available for each week.
Level the schedule using 15 workers. This projects critical path is 47 wks.
Hint: First level the schedule considering unlimited worker then adjust the
schedule for maximum 15 labour

Activity H I J K L M N O P Q R S T U V W X Y Z
Duration 10 8 6 4 2 4 4 5 5 5 5 3 3 1 5 2 3 8 6
Wks

Labour req 5 4 8 2 6 3 2 5 6 2 0 0 0 9 14 6 6 14 5
(wkly)

IPA - H H J J J I I I,J L M N Q O,S K,R U V P,W,X YT

Schedule and corresponding worker load
considering unlimited resource
Schedule and corresponding worker load
maximum availability of 15 workers at a time
Schedule and corresponding worker load
maximum availability of 14 workers at a time
 Lecture 2b

Time Cost Relationship:

Crashing Program

Dr Anis Rahman

G09 1.19
a.rahman@griffith.edu.au
CPM
• Any given activity is assumed to be performed at normal
work pace and finished at normal time – this is the
normal point;
• Time associated with the normal work pace Tn– how long
the activity will take under normal working condition;
• Associated with the normal time is the normal cost – cost
to complete the activity under normal working condition
• Normal working pace is assumed to be the most efficient
and thus least costly;
• Extending this activity beyond this point would not saving
anything but might well increased the cost.
Time cost relationship
A mathematical procedure is used in CPM for
estimating the trade off between project
duration and cost.
 It assumes estimated time duration can be
reduced by applying additional resources to
particular key activities.
 It also assumes that time to perform an
activity in the project is variable depending on
the effort or resources applied to it
 Crashing Program

Reducing the project completion date by

evaluating the plan and re-scheduling it with
network logic or resource allocation changes.

•To reduce the time to complete an activity needs more

resources in the form of additional personnel and/or
overtime.
•The more resources are applied- more the project
duration shortened with the higher cost.
•When maximum resources are applied to complete
an activity in the shortest possible time, the activity is
called crashed
Cost slope concept

 Can be used to determine the most efficient way of

shortening a project duration
 Represents the time cost relationship or the
marginal trade-off of cost to time for any activity.
 Every activity has its own unique time-cost
relationship
 Relationship can be linear, curvilinear or step
function because the shape of the actual time-cost
is often not known.
 A simple linear relationship is assumed
 Cc − C n
Cost slope =
Tc − Tn

crash
Cc Slower than the normal
Cost $

Cn
normal

Tc Tn

Activity duration
 The objective of each stage in crashing a schedule is to work
on one (or more) activities to produce a feasible shorter schedule
for the minimum increased direct cost per unit.
$5000

Cost
Steps:
1. List all activities on the critical path
2. Determine the cost slope of each task $3200
3. Select the activity with potential
(time savings for least cost: low slope)
4. Determine new critical path(s) if any
4 10
5. Re-evaluate the network Time
6. Repeat the above for another activity.
cost slope= $300/day
Exercise

 Shorten the duration of the following project using crashing

program (all cost values in $Ks and time values in weeks)

Activity Normal Crash Cost

slope
Time Cost $ Time Cost $ C
(days) (days) F
A
D
A 9 10 6 16
B 8 9 5 18 G
B
C 5 7 4 8
E
D 8 9 6 19
E 7 7 3 15
F 5 5 5 5
G 5 8 2 23
Name

Student Number

Shorten the duration of the following project using crashing

program
Activity Normal Crash IPA
Time (days)Cost $ Time (days)Cost $

A 4 210 3 280 -
B 9 400 6 640 -
C 6 500 4 600 A
D 9 540 7 600 A
E 4 500 1 1100 B,C
F 5 150 4 240 B, C
G 3 150 3 150 E
H 7 600 6 750 D,F
Monitoring and Control
 Project Monitoring is the set of
procedures and management practices
used to collect information about the
performance achieved or forecasted in a
project and the developing organisation,
based on a set of performance metrics.
 Performance Analysis: The process of
determining performance variances based
on monitored or forecasted performance.
Monitoring and Control
 Project control is
 the adjustment of the project activities to meet its goals by
assessing the performance of the project, analysing the
causes of performance problems,
 designing changes to address problems that are determined
to need attentions, and implementing those changes through
control actions.
 Project control is distinguished from project planning in
two important ways:
1) project control yields a set of designs, decisions, and actions,
whereas project planning yields a design, and
2) project control is a real time process during the
implementation, not before the implementation begins
(MITCE, 06)
Components of Effective Monitoring

 Representative performance metrics (established at planning phase);

 Cost & schedule milestones should be well-defined and clearly
approved/rejected;
 Reporting schedule
 Financial importance of activity
 Activity criticality
 Rate of work
 Difficulty of work;
 Management Scheme organized for honestly and accurately identifying
and reporting performance;
 Involvement of responsible and knowledgeable people in the reporting
scheme
 Project Reviews (walkthrough's & inspections)
 Project Audits.
MONITORING USING BAR CHARTS

Time-now date

On schedule

Behind schedule

Activity progress
 MONITORING USING NETWORKS
(4)
RD = 2
5 8 13

FF TF
TIME-NOW is Workday 10
C
5 13

(2)
RD = 7
(3) (3)
0 5 5 5 7 12 13 7 20
FF TF FF TF FF TF
Will start as soon as
A B E
5 20
D is complete
0 6 13 13

(4)
ss 1 6 4 10

FF TF
D
9 13

RD = 4 Hasn’t started but will start today

RD = remaining duration
 MONITORING USING KEY TRENDS
Work remaining

0%

Planned progress
Required progress
65% r o gress
tu a lp
Ac
Time
100%
Jan Aug

• Identify critical and near critical activities

• Identify milestone dates for completion on a percentage basis %50, %60%, ...
• Monitoring to be undertaken on a regular basis
• Decisions to be made by management to bring the project back in line with the
program.
 Exercise (Monitoring)
 A project involves 4 tasks A, B, C, D, as shown in the figure. Task D has
two immediate predecessors, ie tasks B and task C, and task B has one
predecessor task A . it was plan that task A and task B would take 10 days
to complete, while task C was planned to take 18 days. The start date for
task A was 1 June (early morning), while task C was started on the 3rd June
(early morning). At the end of 6th June, the project manager assessed the
progress of the project and believed that 50% of task A and 16% of task C
had been completed. If the work rate for tasks remained the same and the
work continued during weekends and public holidays, at this stage,
1. What was the expected early start date for task B?
2. What was the expected early start date for task D?

B
A
D

C
 Solution Example 1
H Activity t0 tm tp te S V

2 6 I
A A 6 14 16 13 1.67 2.79
G B 1 2 3 2 0.33 0.11
E
C 3 12 21 12 3.00 9.00
7 8
1
K D 2 5 8 5 1.00 1.00
D
E 6 15 30 16 4.00 16.00
J
B F 3 4 5 4 0.33 0.11
3 4 5 G 5 7 9 7 0.67 0.45
C F
H 4 11 12 10 1.33 1.77
I 1 5 9 5 1.33 1.77
Paths Te V
J 1 5 9 5 1.33 1.77
1-2-6-8 28 6.34 K 1 4 7 4 1.00 1.00
178 20 17

12578 29 6

14578 18 3.89 Critical path

134578 27 12
Solution

 Find the z

z = (Tx-Te)/S = (27-29)/√6 =-0.81

Find the probability from the PD table

Ans: 20.90%
Normal Distributions

 All you need is the mean (µ ) and SD (σ ).

 The 68-95-99 rule suggests that all normal distributions share
common properties.
 One standard deviation away from the mean in either direction (red) includes
about 68% of the data values. Another standard deviation out on both sides
includes about 27% more of the data (green). The third standard deviation out
adds another 4% of the data (blue).
 In fact all normal distributions are the same if we measure in units of
size (σ ) about the mean (µ ) as centre.
 Changing or conversion to these units called Standardising
The Standard ‘Z’ Scores

 In statistics, a standard score is a dimensionless

quantity derived by subtracting the population mean
from an individual raw score and then dividing the
difference by the population standard deviation. This
conversion process is called standardizing or
normalizing.
 Standardscores are also called z-values, z-scores,
normal scores, and standardised variables.
The Standard ‘Z’ Scores

 Computed by subtracting the mean (µ ) from

the score and dividing by the standard
deviation (σ )
For sample Z-score: X −X
Z=
s

For population Z-score: X −µ

Z=
σ
Name
St Number

 Shorten the duration of the following project using crashing

program (all cost values in $Ks and time values in weeks)

Activity Normal Crash IPA

Time Cost $ Time Cost $ D

(days) (days)
H
A 4 210 3 280 -
A

B 9 400 6 640 - C
F
C 6 500 4 600 A
B G
D 9 540 7 600 A
E
E 4 500 1 1100 B,C

F 5 150 4 240 B, C

G 3 150 3 150 E

H 7 600 6 750 D,F