Scheduling

Scheduling Techniques in Projects
Dr. J. Uma Maheswari

Department of Civil Engineering
Indian Institute of Technology, Delhi
Lecture - 01
Introduction to Project
Welcome to all of you. So, this is the 1st week 1st lecture on the course Scheduling
Techniques in Projects. So, the 1st course is primarily introduction to what is
Scheduling.
(Refer Slide Time: 00:33)
So, as we will see in this particular class there are four topics which I will be covering
here. First is what is project, what is construction projects, project management and what
are the three main themes in the project management which is very critical for any
project for success of any project I would say; planning, scheduling and control. In there
we are going to only focus on a very small topic which is only on scheduling and how to
apply the different available scheduling techniques, under what situations we have to
apply.
So, that is what we are going to learn in this entire course. Since the scope is focused on
scheduling, so, we will also be seeing on an overview on what is scheduling, what are the
different types of scheduling availability and so on and then, work breakdown structure.
So, with that we will finish for today’s class.
So, now what is a Project? Some of these things may be little boring for you because this
is like a basic backbone of any construction management student, but still I thought I will
just freshen it up and then, so that you will have a little idea on and you will come to a
same common platform. So, what is a Project? So, a project there are so many definitions
all over the place and if you look at the famous PMBoK definition on project, it has a
collective several stages. One unique words which I would like to underline here or
every project is unique. For example, manufacturing a stapler. So 1000 staplers if you are
manufacturing that is not called as a project, but the initial arrangement in which you are
trying to figure out and then do primarily in how to get out a stapler, we can call it as a
project.
So, every project is primarily unique. In that it requires doing something different from
what was done previously. A project must have a specified objective to be completed
with certain specifications. So, what is it I wanted to do? Maybe I want to build a house,.
That is a specific objective I have there and certain specifications also I want I want a
very lavish bungalow set up or I want a very simple home and with minimum cost into it.
So, there is some specifications on guidelines into how I want to achieve my objective
and there should be a specific start and a specific end date for the whole project. Maybe I
want to start the house construction maybe in the next one month or so and I would like
to finish it in 6 months. There may be a specific plan in everyone’s mind when you want
to perceive any project and you should also consume resources. Resources can be in
multiple forms; manpower materials, machinery, money also can sometimes be a major
resource component when you are working on a project.
Now, if you look at the projects, so what can all can be called as a project? Projects can
have a varied range. It can range from simple repair works or a medium type of works
taken up by most public-private sector parties or the complex construction projects like
your roadways, railways, airways you can take. So, all these can be called as a project.
So, whatever is you want to classify it should have the above defined objectives, then
they are all called as projects. We may also have urgent projects, secret mission projects.
So, the way we operate all these projects will vary from one to another, but everything
can be called as a project.
Now, what is a construction specific? Let us go little into construction. So far I gave a
very broad definition. Now we will see what is construction. There are some typical
characteristic features of any construction industry. For example, it is a largest industry
in the world that nobody can deny. We are the largest employers in manpower compared
to any other sector next to agriculture.
Most projects exhibit time delays, cost overruns or conflicts among stakeholders, every
time no difference of opinion or maybe they have understood the project in a different
way. So, there are some conflicts at each and any stage of the project with primarily with
the stakeholders and most of the projects are always boiled with uncertainty, complexity
because many of the reasons can be narrowed down to lack of clear understanding in
what the client requirements are, apart from the real complex nature of the project itself.
So, these are the characteristic features of any construction industry you can talk about.
Apart from there are other issues like environment situations, then unskilled labors in the
in the industry, all these enhance the complexity and the problems which we face in
construction industry.
Now how does a construction project start? So, the construction project generally starts
with a project owners desired to do or perform something it can be create and for
example, if it is a government project, it can be even created something like an Airport
project or something for if it is a private party it can be set up in apartment projects. So,
it can vary from people to people, but there is primarily a desire to do something. And
success of the project lies in how best the typical two parties apart from the client or your
designer and the contractor and how they have understood or perceived completely the
owners’ desire and they have implemented in their own design or in construction. There
lies the complete success.
So, if you look at project management in construction industry or what we call it as a

construction project management, it is primarily a professional practice of planning,
scheduling and controlling a finite number of resources to achieve a set of goals. So, very
simple definition we have, but when you start working on your project you will
understand this is really complex in a construction environment.
So, with this background now we will see where we are, right. Now, let us just give a
glimpse on why is it so complex right now, ok. So, project management in the past 50-60
years and the project management done recently in the last 10 years maybe if you want
to look at. So, I just wanted to just list out what all the projects have happened,
construction projects has happened. If you look at the traditional project I have given a
name called Traditional and Modern. It does not mean the management techniques were
traditional and modern, it only I wanted to say the timeline with which these projects
have happened.
So, if you look at few examples maybe India Gate, Qutub Minar I am just taking
examples in and around Delhi only. India Gate, Qutub Minar, Red Fort, Lotus Temple,
Rashtrapati Bhavan, so you can see so many examples which had happened in the last 50
years back, ok. The current projects if you look at Akshardham temple, Delhi Indira
Gandhi International Airport, Delhi Metro, Yamuna Expressway very successful projects
way I can list out on the other extreme also. So, how was the project management differ
in those days compared to now? In those days there was a rigid environment. They took
so many years for completion maybe they did not even have a proper mechanism to look
at time overruns, cost overruns impact of all these on to the projects. There was one
deciding authority and the entire team had gone ahead in the same way.
Now, we have many challenging environment. If you want to even the owner starts
asking how much, why are you delaying the project, I want the project in one year. So,
now the knowledge is like spread across a whole community and so the environment is
very challenging and everybody wants to build something unique compared to others and
they wanted to be in the name fame. So, all these have created the challenge in the
environment. So, there is a changing trend set now and owners also want more and more
in these days. In the earlier days people always have worked in silos so which implies
somebody has told me what I have to do. So, I am only very focused with that alone. I
am not worried about what x y or z in my team are doing at all in the project.
But right now people are very curious in knowing what others are doing, so that they can
also share ideas, share their tasks or they can get some you know inputs from them, they
can also share their outputs to others. So, there is an integrated project delivery sort of an
environment which is still existing in these days I would say. So, the one advantage I
would call in the traditional project management methods are the entire team was sitting
in one place close to the project site or in and around in the project site till the project
was completely done which is primarily we call as colocation, but right now the entire
team is most of the projects you will see this is a common scenario. They are spread
across the entire globe and they primarily work, try to operate using clouds, share their
models using clouds or they send through emails whatever mechanism.
So, primarily the face to face communication, the real constraints which are faced by the
people are not really understood by the team and there is a lot of issues there, but that is
how the today’s environment has to go because with the technology fast in pace people
do not have the time to travel to one place to another, sit there and do nothing other than
focus on the particular project because they want to do multiple projects at a one time.
So, and the list goes on and, but just comparing on what happened 50 years they were
also successful projects in those days and why still project management is really getting
into very challenging mode. I just thought I will just put down all these points.
So, primarily this is what is a current scenario right now. The owners want more and
more. The scheduling technique was available 50-60 years back were not enough at all
right now. So, we wanted more techniques, more solutions, more challenges, more
innovations. Obviously we have to keep the project management to the latest trend as
with the market.
So let us start looking. As I told you the scope of this entire course is towards scheduling.
So, let us move more on into the scheduling part. With the last few slides you would
have understood the planning, scheduling, and controller are the three triangle of any
construction projects. So, this is a PMBoK Five process groups. It is a very simple
diagram. I thought I will just showcase this to continue with my team. First you have an
initiating process and then there is a planning process, then execution and control. If you
see here planning, execution, control itself is a cyclic process and the execution, control
is also a cyclic process. You would see that there is a double arrow mark for the same
and then there is a closing process.
So, this primarily emerged from the PDCA cycle on cyclic improvements and
innovation. So, that is the main idea behind this. So, planning process it is like it is
dominant all through even until control phases. Planning is like a first step and it is a
continuous step which runs all along the construction till execution and stops. Then next
comes execution. Execution primarily involves all facets of your construction
management starting from time management, cost management then you have contract
management, safety management, quality management, human resource management,
communication management. It has so many issues to coordinate in the execution
process and the control processes are primarily in terms of how best you can manage
your time or cost and bring it back to control suppose if it has slipped beyond your
control or something. So, that is primarily your control aspect.
So, as I have explained here, so scheduling is one aspect of the execution process which
we will see little more in detail. So, now moving on into little idea on what is planning,
scheduling and control. Obviously all three has to go together, but we are only going to
focus on scheduling, but little idea on what is planning and scheduling it is essential for
us to move ahead. So, what is planning I have drawn a small cartoon over the top.
So, primarily it is a mental map with which a person has in his mind. So, which is like if
I have to do this say everybody even if you want to go out for a party you think about
what to do what dress to wear the when should I start at what time should I go there what
is the mode of transport for me. So, there is sort of a mental map everybody has in their
mind and if you will see this in day to day activities in your in your life.
So, planning is generally an activity of the mind and planning, there is a definition given.
It is a creative and demanding mental activity of working out, what has to be done, how,
when, by whom and with what resources. So, primarily it is doing the job in mind. You
do so many iterations in some literature people also say it is a mental iteration which is
not shared, laid down or drafted out, but the iteration goes on. Is this feasible or that is
not possible because I have to reach in time, I cannot use this.
So, that mental mapping is always worked on in somebody’s mind that is primarily a
planning process. I have brought it out for you the planning process is of paramount
importance in success of construction projects and there is statistics that was quoted in
the book. One study has showed exceptional planning effort can yield as much as 40
percent cost savings over a reasonable planning. In the same study, it was also shown
poor planning can create as much as 50 percent of overruns.
So, if you have a proper planning obviously you are going to see much benefits in the
downtime of your execution part and if you have not done a proper planning, obviously
you are going to end up in time cost overruns, whatever, then you are going to have a
failed project in a way. There are many approaches to planning and each person can
follow their own convenient level, but the process as a whole is not very scientific or it is
not even systematic and as it is practiced in these days and exceptional planning also
involves lot of documentation, or assumptions, analysis made etc in order to finally
arrive at the identification or list of work activities.
And the plan also comprises of a number of interlinked plans. For example, quality plan,
safety plan, human resources plan. You may have so many plans and sometimes it is
better to have interlinked plans to for an integrated approach.
So, next so first we have seen planning, the next is scheduling. It is determination of the
timing and sequence of operations in the project and their assembly to give an overall
completion time. So, some people they also use like a formula. They say scheduling
equal to planning plus time, but I do not want to put it exactly in a formula mode. So, it
is primarily adding duration in order to uncover all your flaws in the plans in order to
have a systematic fare through of your planning process is primarily called Scheduling.
So, scheduling is fairly systematic and scientific and can also uncover any flaws in your
plans and the project schedule may mean different to different stakeholders, but
generally it is used as a tool to manage your time aspects. So, whenever you talk about
scheduling, it is only the time management I am going to talk about in this. So,
scheduling is truly an output of the planning process which is generally an art and it is
never done by computers and it is only done by people. Do not think I have a list of
activities, relationships and durations. I am just putting it onto the computers and the
computer gives me an output scheduling is done. It is never done by computers and
human intervention is always required and you need to know what is the flow, how is it
have I done a mistake somewhere. So, scheduling is obviously done by a good scheduler.
Next Project Control. So, the minute control happens I am only going to talk about
planning your project schedules, scheduling process and you are controlling the project
schedules. I am only going to talk about time management in the planning scheduling
and control. So, when I am talking about time controls, so construction projects rarely go
as per the plan. Either it goes ahead or generally it goes as per the plan and sometimes it
also goes beyond the time durations .
So, the project control it is a continuous process which has several steps and the few
steps are measuring the work progress. So, primarily you have to show if I am working
on an activity which is taking like 10 days of duration. On the 5th day you want to know
how much of the work has been done progressed and how much is supposed to be
executed, how much has been really executed on ground, what was a reason, what is a
root cause for the delay accelerated work etc. So, that measurement has to be done first
and then it has to be compared with the baseline schedule or budget if you want both and
then find any deviations.
If there is no deviations, then fine you can go ahead with your processes. Suppose if the
project has taken too much of a delay or it has not even started on the on maybe on 5th
day or 6th day and you are just waiting for something, some approval or some materials
to come and you are just waiting, then you have to analyze what is a reason for the delay,
how to uncover those delays or what is the corrective action you wanted to do and if you
are starting on after 5th day, then how much you can finish in the remaining 5 days, do
you want to have more resources and how do you want to finish it in the 10 days plan.
So, all these are primarily a controlled process.
So, control process is nothing, but it is not just measuring what has happened and what is
to be plan, it also talks about corrective actions measuring and so on. So, it is a very
broad theme. I would say on the control. So, project control generally refers to time and
cost control only. Other controls are also available like quality, safety, etc. which has to
be dealt in a very different manner. Only time cost control is really easy to monitor and
execute in the sites and next is Schedule Compression.
So, whenever if your schedule has slipped, then you may have to use some techniques
which is called Schedule Compression. Compress your schedule or you do fast tracking,
time cost tradeoff, another issues fall in place there.
So, now let us move on into Scheduling again. I am just bringing the five process groups
which I have shown you earlier. So, I have planning, in the execution part I am going to
talk only about the scheduling part and next is a control processes. So, since I told I am
only focusing on scheduling, so we are going to see on how to do the little partial plans
on building up a schedule and then we will stop with the scheduling techniques available.
So, now what are the benefits of scheduling? Why should I do scheduling? There are so
many benefits available and the list like goes on, on and on. Let me talk about from a
planning point of view. So, there are several views I have highlighted in different forms.
I should know how much of resource has to be assigned in the project, how many
resources has to be consumed for activity A,B,C etc. So, only when I know this, I can
always build my schedule. So, primarily this helps you to assess or assign the resource
utilization. Anyway while we are discussing on how to derive at activity durations, we
will also see little on the resource combinations how it is really playing a major role in
arriving at your durations. Next is to predict project completion time. Only when you are
putting individual activity time durations and then you are having a very large network
diagram, then you will really come to know what is the time completion for my project.
Is it one year, six months and so on that curiosity everybody will have on the project
management side. The next not only on the total project completion time and somebody
really wants to also know what is the individual milestones completion time.
Suppose if I want to build a very big lecture hall complex or something. When will the
classrooms be ready, when will the auditoriums be ready, when will the labs be ready, so
there are several milestones in the whole project and somebody wants to know when are
these milestones can be achieved. So, individual milestones also you will come to know
when will they be achieved. The next is to avoid conflicts or clashes in the work
sequence as we have seen the in the scheduling itself. When we are assigning your plans
with the time duration, you also try to uncover all the flaws or the mistakes you do in
your planning process itself.
Some unachievable plans whatever you have methodologies, you have all get uncovered
when you are putting it on in your schedules. So, it can also avoid conflicts clashes later
on. If you do not have a proper scheduling, it may it may later on pop up when the
excavation when the execution is going on. So, it can also help you to avoid conflicts or
clashes.
The next important thing is I can also know what is my cash outflow and what is my
cash inflow, when all my cash is spent on various activities and when will I get the
payment, how much work I can I can complete till that date, so that I can get the
payment back. So, I can easily plan or calculate or anticipate my cash outflows and
inflows. So, this scheduling document is used for multiple purposes. The next as we have
seen in the control slides. So, to monitor and control the project progress obviously how
much has happened and how much has to be happen and there is the gap between the as
planned and the as execution. Primarily it helps you to monitor to estimate the impact of
changes. Suppose a owner in the midway of an execution says certainly that he does not
want X or Y part in the whole project and he wants to be changed something, then you
should also you can also easily come up or arrive at what is the change because of a
small segment modified or a new segment added up in. So, easily you can estimate the
impact on changes on a projects. It also serves as a record in terms of delays in the work.
So, sometimes when the delays are there, is it an owner related delays or contractor
related delays and is it like shared, is it on the critical path, is it on non-critical path. It
also helps you to record.
To settle or verify disputes in case of delays which is what I said. So, suppose if the
owner and contractor both have a mutual misunderstanding saying that this is not from
their part the delay has happened or the other party has to take care of all the penalty,
then this document on scheduling, the initial scheduling helps in a way to settle all the
disputes and verify the case properly and the list goes on to the benefits of scheduling.
Next Types of Schedules. There are so many types of schedules. Anyway when we see
the duration estimation, we will also have little idea on how that individual activity
duration have emerged and also how these names have come in. We will see little more
at that time.
So, types of schedules. there are broadly three types in the arbitrary basis. What happens
is the owner just indicates a project start and end dates. He says the project feasibility
time of start will be next after one month and the end date will be after 6 months. So, he
just arbitrary indicates. So, the real practical constraints are not at all analyzed in these
arbitrary basis because it comes as an urgent mission or it comes as a compulsion or
pressure on the person. So, these type of schedules we call as a schedules are drained out
of a pressure from the clients.
The next is intuitive basis. So, these are all prepared without using any formal scheduling
techniques and very effective when the contractor has extensive experience on projects.
Suppose if the contractor has worked on a project and he is supposed to work on a
similar project in another 5 or 6 years. With his experience vast experience he had in the
previous project he can always use his intuitive basis to in order to fill up the entire
schedule. So, those type of schedules we call it as an intuitive basis.
The next scientific basis. Here, formal scheduling techniques are used for analysis and
for schedules and the contractor uses methods for scheduling and he executes it. So,
fairly a scientific basis which is what we are going to see right now and here there are
three types of schedules which researchers have identified. There are again many more
types, but we are only going to see three. One is logic driven schedule; the other one is
resource driven schedule and the third one is information driven schedule.
So, what is a logic driven schedule? As the name implies the logic of the construction
derives the schedule. For example, after foundation then only you can do your columns
or beams and slabs. So, that logic primarily derives your schedule and that is primarily
called a logic driven.
The next is resource driven. Suppose I am having two or more crews for doing a
particular work. So, based on the resource crews I can always do the work in parallel and
are based on less number of our scarce resources that I have. I may have to sequentially
execute the project. For example, there are three activities which requires the same
resource. So, all these three activities will be done one after the other if there is only one
crew with me, but if I have more crews, maybe two crews, then two activities can be
done simultaneously or if I have three crews, then all the three activities can be done
together. So, those are primarily the resource driven schedules and the last is information
driven schedule. We have to define many more terminology to arrive at the word called
Information.
So, information is nothing, but height of the structure or width of the beam reinforcement
spacing cover to the reinforcement. These are all called information in a project. So, this
is these information drive the schedules. So, primarily this is part of your design phase of
the execution. I would say we have several design phase examples to cover on
information driven schedule, but never assume that it cannot be applied in construction
as well.
If you have information driven data in construction, it can be still applied in construction
phase. So, in that way the first two logic driven resource driven I may be planning to
share more examples from construction sector and information driven primarily from
design phase of the projects. That’s how I am planning to go ahead and these three
schedules we will be seeing in elaborate.
Now, let us come to the PMBoK time management flow diagram. So, now actually we
have moved on in to what is scheduling. So, as I have been telling you scheduling is
nothing, but planning plus time. So, planning there is lot of homework to be done as to
what is the scope of the work, what all has to be done, what is the critical milestones in
the entire project, everything is decided in the planning stage itself and then you add
duration sequence to the project sequencing with the help of relationship between the
different activities and then you move ahead. So, if you look at the time management
flow diagram, there are few steps given in the PMBoK. One is define activities. Why do
we need activities and what is an activity definition of activity and etc. we will see in the
next lecture, but since the name has come right now I will just give a small example.
An activity also has a similar setup as a project. In a very simpler layman language if you
want I can tell. So, project and activity both have similar terminologies in the sense both
are having a definite start, definite end, both have a specific objective, both the case they
consume resources, then why project suppose if you want to think about an auditorium
project it is very difficult for you to monitor and six month project in one go.
So, these activities are nothing, but the breakdown of the project into smaller groups, so
that you are able to perceive the project and you are also able to control, you are also
able to imagine all the assumptions and then monitor the whole project. In that way we
call it as activities and hence we have excavation, foundation, column footings,
shuttering, de-shuttering, we have so many activities from that point of view.
Then the next is sequence. The relationship between each activities, once you determine
that, then you can easily get the sequence whether I am having all independent activities
or dependent activities, what is the sequence between those activities can be really seen.
Then the next is Estimation. So, primarily I need to have as I told we are seeing on
scheduling. So, it is time aspects.
So, time aspects can come from two forms either you fix the duration and decide your
resources or you decide your resources and then fix your durations. So, both has to go in
parallel. That is where I have written it like this. So, one is called duration driven
schedules or it is resource driven schedules whatever. If I have 5 crores, my duration of
the activity will be different. If I have to finish my activity in 10 days, then what is the
crew size I have to go ahead? So, that is primarily the two differences we have and then
you develop your schedule mechanism.
So, this is a broad flow or diagram which we have for Time management.
So, now first task define activities in order to arrive at a list of activities, there are some
techniques available. The very famous technique in construction is your work breakdown
structure. Some people say this part of a planning tip process, also some people say it is a
part of a scheduling process to arrive at a list of activities. So, I thought I will cover this
for today and then in the next lecture we will see the definitions of activities,
relationships, durations for each activity and so on in detail. So, Work Breakdown
Structure the famous short form is WBS. So, what is this WBS? It is a hierarchical
representation outlining the structure of work for a project.
So, the smaller and smaller groups of work until the lowest manageable level often it is
called activities. That is primarily called as a work breakdown structure. In general, it is
an iterative process. Why? Because as and when you start developing a WBS and the
scope of the work is also determine along with the along with the WBS generation, so at
10 percent start of the project you have very little information on the project, then after
20 percent start of the project you have some more information coming in, some more
players coming in, some more scope of the work gets clarified, then you may have more
data coming in. So, then you may have to try, modify, do this and this. So, this keeps on
evolving and hence it is generally termed as iterative in nature. And WBS can be used as
an instrument for scheduling, personal assignments, which implies human resource
assignments, resource allocations and measuring the project progress etc. So, it is also
used in a way for doing that. And the purpose of WBS is to demonstrate clearly to all
parties involved in the project, how the each task is related to the whole project in terms
of budget, schedule, performance, responsibility of who is going to execute the work or
the physical assets of the whole project.

For example I have given an example. This is one activity example. So, at the end of a
WBS somebody has to build up a chart like this, not necessarily in the same form, but
these are the information which goes in into that activity. Activity ID, this is primarily a
housing project and the first activity in the housing project is called Line plan. So, what
does it Line plan all about? So, description is line plan. Assumptions made. What is the
assumption? The customer is supposed to be available till the architecture drawings are
finalized. That is an assumption made for the line plan execution. Time required
approximately one week. Resources how many people you want? One architect is
enough for consultation and even that person need not be available permanently, maybe
one hour per day is enough, maximum of 6 hours is more than sufficient for the whole
work and the discussions can happen through telephone, email, discussions and what not.
So, what is the cost consumed for this activity? 2500 rupees we have to pay for the
architects per day. That is the whole cost part. And who is responsible to execute this
line plan? It is a Project Manager. So, who is in-charge for executing this work? It is a
Project manager.
So, like this we have to derive for each and every activity that naturally comes with all
the assumptions you are putting in and you are actually deriving and decomposing your
project into various levels.

Now let us see an example also. This is a simple example of developing a construction
management laboratory. Actually this was a live example which happened in our
premises, but I have modified the example just to showcase the problem as such. So, this
is a room. So, the room was not functional for few years and this was supposed to be
taken up by the construction management group in setting up their laboratory. So,
actually, what is the initial scope of the whole work? This has to be developed for a
teaching laboratory under research laboratory. So, teaching means teaching for
computational software for students, PG students and also as a research laboratory for
the PhD students and MTech students. That was a whole scope of the work, and the
dimensions of the project are given. So, there is a door that cannot be replaced or shifted.
The doors places is fixed, dimensions of the room are given. So, with this the project
started.
So now what happened? So, the critical requirements of the lab are listed down. So,
somebody started noting down what are all required to be done with the given scope. So,
it has to be functional. The functional features of the space of the rectangle space is
supposed to be served as a computational lab function. So, that has to be set up. So, the
room space, room interiors, room facilities, hardware, software, prototype models. So, it
is an outline and two or three meetings happened and some lists came from the from the
user groups. User groups were the student groups and the faculty groups and some list
came emerge together which I have just written down.
Room space, interior space will be completely renovated because it was very old and did
not have the ambience for students to sit and work and for a class to be conducted. Room
facility - One bookshelf was necessary and people thought there should be a bookshelf or
somebody to refer books and then know work on their projects and research. Projector
for teaching purposes, display boards for a putting notices and so on. One whiteboard, 1
or 2 AC units for so many people and computer sitting in the room they wanted 1 or
maybe even 2 AC units and CCTV cameras for security purposes. Hardware - So, high
end computational facilities maybe in the form of workstations were planned and
proposed. Software for example what these construction management group generally
use. So, MS project, Primavera, Tilos, formwork software, so all these were. So, the list
went on and as and how many where they wanted to implement that many software for
installations.
Then prototype models also because it was supposed to be a construction laboratory.

They also wanted working models, may be a small backhoe or a crane set up showcase to
the students and teach in the class as well. So, the ambience was also planning for a
display models and next is Occupancy Status. So, seating capacity approximately with
the current situation we plan for 10 PhD students and 20 MTech students who work in
the projects should be able to sit very spaciously and without less disturbances they have
to sit and work in their labs.
So, after the purpose was drafted out, room was given after two or three meetings had
happened. The first cut idea on what is a requirement for the lab emerged, then what
happened was so this I am highlighting it because I am going to explain this in detail. So,
seating capacity was 10 PhD and 20 Mtechs. Then, the WBS structure started evolving
which should define the scope of the whole work.
Keeping these client the requirements of the students and the users. So, WBS I thought I
will just put it for you. So, I arranged in under three categories; room related, students
related, faculty related. Now, WBS as I told it is a hierarchical tree form and how many
ever levels you wanted to go, you can still go and finish. There is no harm and there is no
constraints as hard and fast rule as to how many levels you have to go. Is it three levels
or four levels or what is the lowest level of the base of WBS? There is no hard and fast
rule on anything and how much you are comfortable in building up or in the hierarchical
breakdown, you can still go ahead with that.
So, I have purposely brought a non construction example in order to showcase WBS can
be done in any form. It is not necessarily that project, then system, component, sub
components, you need not go in some hierarchy. You can go in anyway, but the work the
next level or your lower level of WBS should always be when you are adding up, you
should always supplement your upper level of the WBS that you should keep in mind.
For example, under the item number one whatever I have highlighted it should all
completely describe the item called room related, so that you should ensure that it is
happening and you can go in any order or level you wanted. So, from that angle, so what
do you want in the room related. So, room related tasks were all envisioned. So,
primarily it is Plaster of Paris or POP. So, the room interior was very bad. So, there
should be POP, whitewashing, painting was essential. False ceiling was required in order
to avoid too much of current electricity issues. So, false ceiling was proposed. Flooring
was also proposed to be modified, then AC was planned. So, AC units and all those
lights, fans etc., fire protection systems, CCTV camera, biometric attendance, all these
for security purposes in the whole room because so many of inventory were coming into
the room space. So, all these were proposed in the particular room. Students related
items. So, seating space. First is they need a seating space, a table and a chair to sit and
work. Arrangement of tables and chairs was also an essential criteria there. Then
computer, computer accessories, printer scanner that is optional and UPS was there.
Software- there are two categories in software. One is a class oriented software or the
basic software I would say and the research oriented software which only few students
will be using in the particular lab that was also planned for, then storage space for books
or other documents for the students to come and work refer something, then they need
not carry it all the way for their lunch and then come back. They can leave it in the same
lab and then go. So, they needed a small storage space for storing their daily routine
items and documents which they were working on. Network connections, electricity
connections, all these were primarily students related activities.
Then, faculty related, because it is supposed to be a teaching environment. So, the other
user who will be coming into the lab is a faculty. What are their items? So, primarily
projector because teaching has to be done. Display boards, notice boards for announcing
announcements and instructions. Display models also for teaching in the class to aid in
teaching in the class. Storage space for user manuals, books, records thesis only for
faculty to use and so on. So, these are the various criteria and with this I would like to
say the scope of the work has really you know been identified. This is a stage 2 or level 2
of the items which were drafted. This is how the WBS starts to evolved.
Now, as I highlighted in the previous place on the seatings space, tables and chairs. Now,
I will also highlight to show little more details. So, earlier I highlighted on 10 students
PhD students and 20 Mtech students will be sitting in the lab or using the lab. In the
seating space, primarily the arrangement matters as to how should the table size, chair
positions, so that you can give accommodation to all these people.
So, the initial proposed layout came out like this. Because research scholars wanted
privacy from others in the sense whatever they wanted to work, they wanted a separate
place for computer and they wanted a separate place for working and on their notebooks
and documents and so on. So, we formed it little like a y-shaped and structure or it is like
a three cubicle structure. So, on one part you can work on your computer and the other
part you can sit and work on your documents. So, that was planned and hardly 11 people
were allowed and positioned with lot of space and movement in and around these
scholars. So what happened, it did function as a research laboratory. For a teaching
laboratory, we were not able to accommodate more.
So, the next level of arrangement was planned. So, we planned like regular pattern of a
classroom structure. So, there were like 6 tables because more tables could not be
accommodated in. So, we proposed 6 tables and there were shutters. So, this was the
general layout arrangement. Bookshelves were on the last side of the of the structure. So,
this door cannot be modified and the room dimensions are given here. The display
boards we thought we will just hang it up all along the whole walls. Display boards were
planned and there were bookshelves on the backside of the whole lab. In the front side
there was a projector screen plan. The display models were kept on one side of the whole
setup. So, all provisions that we thought of keeping in the lab were all had come in. And
this was planned as a two-seater or maximum a four seater when the class is going on.
So, in that way research scholars were planned as two. So, this was ideally like 12 people
were sitting, 6 tables-12 people were sitting for the research purpose and when the
teaching was going on this was planned to be accommodated with two seats per
computer and hence, it was also functioning as a teaching laboratory. So, this is the next
level of detail plans working out and then the schedule starts coming in as to how should
the table size be, how should the chairs be, should we buy stools or chairs. So, all those
details will go on when once you start executing the project in the next levels and so on. I
thought i will just show you on this now more pictures on this.
So, this was a rough cut draft. So, primarily my projector screen is lying here, display
models were planned to be here and the door is supposed to open on this side for the
people to enter and go and so on.
So actually, one seat for the research people to sit and when the classes were going on
two people were supposed to sit in all these places, per table and one computer they will
share, so that more than 20 people could be seated and had a seating arrangements. And
the problem with the earlier setup was this. In this setup, even if he had a projector on
one side, all students will not be able to be facing the course instructor and they will not
be able to work simultaneously but as in this model they could do. So, in this even the
layout and arrangement were all fine. So, these iterations also happen sometimes maybe
the example is not that explicitly explaining to you what I wanted to say, but still I would
like to say starting from purpose to serve as teaching and research laboratory with the
funds, time constraints and all those. So, primarily the idea has to evolve. So, this was
the earlier meeting, then this was the next meeting and then just for one item seating
arrangements, 2-3 iterations also happened.
So, this is how the entire WBS process goes on. I just wanted to showcase with this
example. This is another view of the same build structure.
Now, coming back to the table chair arrangement alone. So, the computer was planned
on the table. Network and electricity points were plan through the walls and to the seats
and it used to come below this and then for this table also printers cum scanner was
planned to be here, UPS was a small provision for putting in the UPS, storage space. So,
all student selected activities has been accustomed here. This I would say the WBS was
able to capture the complete scope of the project, some cases it does not do because of
the budget, other constraints some cases you may have to modify also and, but in this
case it was able to capture the complete requirements of the students related.
So, just wrapping up for todays lecture we started on introducing what is the project,
project management, definitions on the project and management etc and then we moved
on into planning scheduling and control, we have seen all the three in little detail.
Scheduling, as we have seen much more in depth and as to the benefits of scheduling,
then types of schedules and what are the various ways of doing the scheduling. For
example logic driven, resource driven and the information driven scheduling.
And then what is the sequence of doing a scheduling activity as such. So, first is defining
activities. So, WBS is a very famous step with which you can define the list of activities
without omitting or without missing any information. You can derive the complete list
by using WBS and also WBS serves as a document for scope definition. It also tells you
what is the scope of your entire work, that I showed with iterations. How the WBS can
be evolved with the same example on developing a computational laboratory.
The same example I will use in the next lecture also to show I would say hierarchical
breakdown has moved on into list of activities relationships duration among the activities
scope modifications, how it got incorporated and constraints that face in the projects. So,
let us see little more in the next class along with defining the activities relationships and
so on. So, the next lecture is also supposed to be an introductory lecture on inputs to
Scheduling only before we start the Real Networking diagram.
Thank you.
Lecture – 02
Inputs to Scheduling
This is week 1 lecture 2 on basics in scheduling. Last lecture we have seen primarily the
inputs or the introduction to project management and then we started little on inputs to
scheduling, WBS we started. We will continue and we will see more details in today’s
class. So, today’s class primarily what am I going to cover? So, we will see activities,
definitions, list primarily the list of activities how to arrive at. So, the WBS example we
will again continue for today’s class. Then, how to estimate the activity duration and the
different types of estimations available in theory, practice we will see everything and the
relationships between activities also we will see.

So, activity; what do you mean by an activity? The elements into which a construction
project is subdivided into manageable minor steps are called activities. While explaining
WBS also I told a project at that level it is very difficult to handle or by a project
manager to foresee all assumptions and other issues. So, the project at that level it is very
difficult so, we have to break down into sub elements in order to manage the entire
project. Manage in a sense; plan, schedule, control, whatever. So, that break down the
last level of the hierarchical breakdown is primarily called the activities and then we
arrive at the list of activities. An activity has a same objective as that of a project in terms
of specified goals, in terms of definite beginning and the end, in terms of consuming
resources, in terms of objectives specifications; everything it has a similar meaning. That
is what is this. An activity is a single work step, that has a recognizable beginning and an
end.
Now, what about the number of activities? Is there any guideline? Should I have 100
activities? Should I have 5000 activities and so on. There is no limit, there is no hard and
fast rule. If I am able to manage my activities at 300 numbers then you can go with that,
otherwise you can go with 1000 also when you are comfortable in foreseeing the project
at that level. So, there is no hard and fast rule in any project. But for every activity it is
advisable that you give a description to the activities, primarily the scope of the whole
activity should come in the description along with attributes which can be measurable
and subjective attributes.
Now, coming to activity types, every time you say there are list of activities and
activities. And for example, I am taking an excavation as an example. In excavation,
there are series of steps I would say. And in the series of steps there may be a value
adding portion, there may be a non-value adding portion and so on. Let us see what are
those categories. So, task types; so value adding activity, non-value adding activity, but
there is another if also non value adding, but necessary portions of those non-value
additions.
Value adding activities the direct physical progressing activities maybe it can be a design
activity or an engineering activity in the design phase or it can be just a construction
execution in the construction phase. So, other segments for example, shifting the
equipments, setting up the whole site, setting up your place and getting ready for that
task, then safety audits, safety meets they are not part of the value adding portion. So, the
value addition is set as direct meaning whatever is a physical progress of your design or
construction is primarily called the value adding activity.
What is a non-value adding activity? Breaks in the work, nobody can work for 8 hours
per day and so on. So, they may become fatigue. So, there are breaks, those breaks can
be a very good example for demonstrating a non-value addition. Those activities even if
they are not in the system still there is no harm or there is no loss of the work going on
work progress then they are called the non-value adding activities. Then the non-value
adding activity, but necessary category, what are those? They can be procurement related
activities, administrative activities etc. For example, checking for the stock list in the
godowns, and inspecting with the local markets and then procurement placing an order
for procurement or it can be an audit inspections, quality inspections, safety meetings;
these can be some sort of an procurement or admin activities which are necessary in the
whole goal. Now, with this I will just show a small example on unloading of rebar in the
site. So, we are very curious to know how much non-value adding portions were there
and how much value adding portions were there in the activity.
So, we used a simple diagram called flow process chart to measure all those. So, flow
process chart. This is the flow process chart. These are all planning techniques as part of
your work study which I will cover up at the end of the lecture on other scheduling
techniques. I will give a small introduction to this. But I will just explain how to read this
table first.
So, the description of activity primarily these are the 1 till 16 and it is only a snapshot, I
have several more down in the line. So, this is only a snapshot on the steps for
completing an unloading of rebar in the site. Primarily I have cropped only till unloading
of rebar in the site and this is a time taken in minutes for each of these steps and these
symbols are primarily to show the weights and then the last column is primarily the
number of labors workers required to do all these operations.
So, now I am just going to read out few of the statements or steps here so that you will
understand when we are really executing the work you will not note down all these
unnecessary steps. But you will see what are all the unnecessary steps and you will
always you should try to improve on the value adding portion to enhance or promote
your constructions projects. Truck with the rebar arrives at site. It is just arriving thus we
are starting from there that takes 0 minutes. So, we are not counting previously before
entering the site, only from the site we are starting our time calculations.
Waiting for the engineer to arrive; so, 25 minutes there was a delay on that particular day
for the engineer to come then the visual inspection on the steel was taken down, it took
another 10 minutes, then receipt was given by the engineer another 10 minutes then wait
for the labours or workers to arrive, it took another 20 minutes after the receipt was given
then the phone call was done to those labours and the labors arrived in another 20
minutes time span.
Then your sample was cut and it was sent for testing it was another 10 minutes, then
cutting the sample for further tests by the third party agencies who is doing the
laboratory test. Then receipt by the quality in-charge that the steel is actually in the good
condition or not as per the specifications that took another 10 minutes. Then traveling of
truck till the weighing machine in order to determine the weight of the steel that took 3
minutes, then wait of the truck was noted down it took 0.5 minutes, then feeding details
of the arrived truck in the system for verifying the receipts at weighing machine. How
much was a load which has entered the site, the receipt has to be entered in the register
that took 15 minutes.
Unloading manually; so, when the truck arrives unloading happened with the small test
with the steel test. So, primarily value adding portions are very less here and there.
Unloading it took 60 minutes, then the truck again empty truck has to go back to the
weighing machine to really determine the self-weight of the steel which has entered the
site. So, if you see the number of small-small steps, you will understand where is non-
value adding and then value adding portion. Effort has to be made in order to improve or
promote the value adding portions and try to minimize or avoiding non value adding
portions in the whole operations in construction.

Now, let us move on to the earlier example which we were seeing in the last class. So,
primarily to derive at list of activities, how do we go level 1, level 2 level 3. What is a
breakdown we can go ahead in order to generate the list of activities? So, we just may
continue with the same example and I am just continuing with the same highlighted task
which we have seen for layout, arrangements decision making and all what happened last
class we have discussed.
So, level 1. So, here if you see level 1, level 2 we have. I have shown you in the last
class. Now level 1, level 2 I am still keeping with the same. Only on table alone we are
going to go a little more in depth. Dimensions of the table; so, arriving at that is one
activity. Finishing the table top and what is a make for the table top, vertical panels that
is another activity. Keyboard tray, drawer unit foot rest, so, these are all the different I
would say parts or segments of the table and wherein if you are making to if you are
planning to customize a table inside, then you may have to procure material ask for
carpenter to make it. So, separate specifications also follows along with each of these
activities; wooden table dimensions is 3 foot 7 inch, length breadth was 2 foot and 2 foot
and the height was 2 foot and 6 inch. The finishing on the top and vertical panels was 25
mm thick plain moisture resistant block board and the keyboard tray was planned as 0.9
mm thick and the width dimensions and all are given with a single extension stroke only
one extension was possible in the keyboard tray. The drawer unit, one drawer cum
pedestal cabinet pedestal of 18 inch width was planned for full height of the table was
planned. And, it should be made of laminated MR block board of 20 mm that was
planned for. In the foot rest, the remaining width of the table you are supposed to have a
foot rest in order to give strong support to the whole table also. That was what planned.
Other things we are not we are ignoring right now. Now, moving on to detailed specs; so,
when you are going on into the detailed levels again you have to work out more on the
scheduling in order to really have a clear idea on the whole project. Detail specs I am not
going to read out, but you can understand this is how everybody starts looking at the
projects.

Now, what are the options available for me? Either I go for customized options; like
bringing in a carpenter, negotiating with the carpenter and then making him do it in the
same site or like here I can order. Or here like this I can order from a website and then
make them delivered to my place. So, these two are the customized options and these
were from a photograph I have taken here and these two by the options. Only thing is
that the specs may not match exactly with these two options. But the specs may exactly
match with these two options, because with these options because it was taken from the
students who are going to be the users and from their specs were derived and this will be
exactly implemented in the site.
Now what is the duration based on the availability of 12 numbers table, on decision on
whether to choose this or this, you will get the table accordingly. The duration of the
table depends on what? The decision that you are going to make. Whether I want to use
this, this may take 2 weeks at least to manufacture all these tables or should I go with this
the availability is there within 3-4 days you will get the table in site. Otherwise it may
take 1 month or so if you are planning to order then get it from the stock and then you
are going to have it.
So, accordingly the schedule durations now vary. Which is what I have written. What
about the duration? Duration is now its not known you have to really explore what can
be the duration based on decision making. So now what I am trying to say is each time it
is not that automatically something is happening, there is a system on decision making
when you go down every level and duration is coming out as a result of that, based on
the comforts, conveniences, constraints which are coming into the projects along with
the requirements of the whole project.
Now, what is activity duration? Now we are stopping at still duration so, what is it that I
want, I really wanted to know what is the duration of my individual activities to really
want to know what is the duration of my whole project. So, what is activity duration?
The time required for each activity for its completion is primarily call activity duration.
So, all time-consuming activities are generally assigned a duration. Now how long this
duration of an activity can be? So, for example, if I have an activity which says like 3
months, 5 months and so on. Should you really keep it as same or should you break
down the activity into sub activities, so that you are able to have a smaller duration.
Again, just like how I told in the activities case, there is no hard fast rule here, there is no
thumb rule on how much the duration should be. If you are able to manage at whatever
level you want you can still go ahead, but generally it is advisable to keep it in work days
or working months rather than in working hours. That is the only suggestion you will see
in many textbooks. Now what is a most difficult a time managing activities: estimating
the correct time which will take for completing a project. Define work unit, for example,
it can be setting up a lab or it can be setting excavation work, it can be a reinforcement
concreting activity. If you do a low estimate what happens is: schedule may look like
that we may be finishing earlier, you may be committing saying that I will be finishing
my entire project in 1 year, but as such on-site it may take beyond. If you are doing
really a high estimate you may show that it may be taking so much days and months, but
actually you could have wrapped it up and finished it up fast and you may end up in
paying more overhead cost, lethargic, slow progress of work. So, there are many
problems which will happen.
So, the difficult challenges how you are estimating an activated duration, which is really
challenging. So, what assumptions people use when they are deriving a duration of an
activity? Activities will be performed normally. Suppose when you want to measure the
distance from your hostels to your classrooms, you should not walk in a very fast speed
or you should not walk very slow. So, the both are not considered good when you are
measuring the calculations. You should be walking in a normal pace and then measuring
up what is the time it will take to travel from your hostel to the classrooms.
So which implies any activity should be performed in a normal speed, in a normal

environment and normal situations by a normal worker also. And evaluate each task for
preparatory items are work. So, how much time is a preparatory item requires and how
much is the real value adding portion in the whole activity and use uniform productivity
rates. This I will I will explain you after one or two slides and as I told you this earlier
you always use work days and not hours for measurement.
I am just bringing a small example to showcase how to find the duration, which we call
as a direct method. There are several methods available we are going to see many of the
methods in today’s class. Direct method using total quantity, crew size, and daily
production rate. In this example, I have said there is a crew and the productivity alone. I
have not told what is a crew composition in this particular example.
For example, find the duration of painting activity of 400 meters square, that is the work
content or work unit quantity using a crew of productivity 12 meters square per hour.
Painting activity we generally take measurements in terms of area only. It is not in
volume anyways. So, its primarily meter square and the crew capacity is 12 meter square
per hour. So, what will be the answer, which is 400 by 12. So, approximately 34 hours is
the duration you are going to take for the painting activity if you are using this particular
crew.
Now the next question which comes to your mind is how am I going to get the 400 meter
square and who gives me the value of 12 meter square per hour. 400 meter square
obviously comes from your estimation exercises. I have a drawing. Generally in
construction what happens is I am given all GFC drawings which is good for
construction drawings. From there you have to really do the estimation in getting out all
your quantities. You have to do working out all the quantities and people use generally
central line method or long wall short wall method there are methods available which
you have known earlier. From there these quantities come there and the crew
composition comes from your resource availability and the productivity of the crew it
can come from the productivity constants also.
Sometimes we can also use experience, suppose if you are not comfortable in or you are
just first time a beginner in going ahead with the project, then you can always choose
productivity constants from the IS code. There is also an IS code available IS 7272 and I
just showed a sample of a snapshot of that, but the only one problem is it reliable and
how many assumptions have gone in here.
Now is it reliable? This data has been taken in 74 and there are plans to revise this IS
code. That is another issue. Now how far can I just use the data for working out a project
in 2019. It is very difficult and you do not know how many assumptions have gone in for
making all these works. For example, say x on level, on same ground level and on maybe
on twelfth floor or fifteenth floor, same work, can you measure the same duration? Can
you think the duration of these two activities will be the same? Activity is the same, but
the place at which you are going to work is; obviously, varying. So, do you think you
will get the same duration for both the activities? Obviously, no, which is what I said
under what situations what circumstances, what is the hazardous nature of the whole
activity, if it is not known or it is not clear at all then you cannot use this standards. But
instead of saying I am not clear on how to go ahead with and you have to build it up and
start using your own measurements from the construction sites. That is what is advisable.
Now, what are the estimating techniques people use or people have as a choice when
they are planning for estimating the activity durations. Number 1; expert judgment; there
is no particular order. So, I am just using my own order. Number one is expert judgment.
It is one of the most frequently used methods for estimating the duration of activities in
projects. So, in this method what happens is that the estimator, he realize on his own
expertise and he consider by some historical information and experience with the
previous projects and then he just says excavation activity: this will take 2 months, this
will take 3 months. So, he just uses that expert knowledge and he starts filling up the
duration of the activities. But what happens is: for improved accuracy it is recommended
that if this can be combined with other techniques which has a little scientific issues
behind, so that it can be done with ease and at a faster pace you can finish up the whole
estimation exercise.
Then the next is analogous estimating. So, it requires you to analyze the previous
projects to approximate the length of the current activity. I would have done 2 or 3 sewer
line projects and this is my fourth sewer line project. So, I may have some different sort
of experience in the previous projects in estimating the durations, I can use some portion.
Maybe I can take a completed project or I can take a section of those project and I can
compare that work with my work how much is a deviations and then accordingly I may
use this method for estimating their duration. In this method what happens you can have
a very quick easy estimates for projects or tasks that are not very complicated. In sewer
lines maybe the ground conditions your design maybe a little varying, but construction
more or less it may not vary. So, you can still go ahead because you may have repetitive
types of a construction work.
The next is heuristic estimation. So, heuristic estimation this is based on parameters
derived from past experiences. In here, what happens is it is actually close to your expert
judgment, but in expert judgment they really go with experts opinion and primarily it is
dominated by experts. And, in this heuristic estimation people use a mathematical type of
expression in order to arrive at a formula and they use that expression and that
expression can be also handled by non-experts in this type of a heuristic estimate when
the expression is made available. That is the underline statement here.
The next method is parametric estimating. As a name implies its primarily talks about
parameter which is nothing, but a small last element and then whole WBS structure. For
example, an activity can be design of a beam or a slab and the parameters can be
thickness of the slab, cover to the reinforcement and spacing of the reinforcement. It can
be so many issues. Same way in this parametric estimating, I have two similar projects, I
have done some one project very earlier and the project which I am going to estimate is
right now here in front of me. Just do parametric estimation. Is it only the windows alone
have changed, there I have used wooden windows and here I am going to use an
aluminum windows and another estimation may fall in place for me. So, that type of a
model you can still see in and you can still capture from there and you can compare and
then you can do. So parametric estimating uses the statistical correlation that exists
between a set of historical data and a series of delineator list of other variables,
parameters based on size, footage or other scope related values can be used to produce
time and cost estimates for the related work. The process of creating a formal parametric
model is to collect data from thousands of past projects and from this produce a
regression type of a model. And this technique can be applied to any situation in which
sufficient historical data are not available which is what I have told you earlier.
The next is Delphi technique and Delphi is a common term which construction
management researchers generally use. Delphi is like I take opinion from few people
then once that iteration is over I start brainstorming on the results that I have got in front
of me. The results are coinciding then you can conclude, the results are not coinciding
and, but they are close to converging again you have to do one round of getting expert
opinions. Then the results asking that person x had given an y value, can you please
change or what do you think about it. So, getting concurrences on others values and then
you start getting at an converged convergence. And the problem here is you may also
land up in several iterations in order to arrive at a converged value. Sometimes the first
round itself you have got totally diverged value. Some people has told 2 days, and
somebody has told 2 months. Then it is like too diverged. And then arriving at a
consensus may never happen at all. That is a real drawback in Delphi technique. Enquire
a from several independent experts to estimate the duration of an activity in your project
and then evaluate the results to provide an objective estimate. This method is based on
the assumption that the independent experts are unknown to each other and does their
individual biases will cancel out resulting in an accurate estimate. I am not worried about
what the other what my other expert has told. So, I am worried about what all I think got
the assumptions, constraints and etc and I am giving my value at all and once the first
iteration is complete the panel meets to discuss the results then estimators review their
estimations in the delta column and this process will continue until a consensus is
reached. Normally Delphi technique is used for more highly complex situations for
which there is little historical background. However, the concept works in situations
where a non-biased estimate is needed, because this technique seeks estimates from
multiple participants, It tends to remove the bias and politics that may happen. Rather
than going with only one expert judgment. I am just having assigning a person for giving
his data and so, he will give only his value which may be more close to the actual or
which may be too deviating from the actual. This can eliminate all those errors because I
am taking opinion from several parties. One of the drawback is as I told you is the
amount of time it can take for the panel of experts to reach a consensus.
There are other methods also on estimating. So, these techniques which I would say
starting from expert judgment, analogous estimation, heuristic estimation, parametric
estimation, Delphi technique and so on; I will show you the problems with practical
examples.
Let us take now one practical example, a very simple earth moving operation. This is a
very classical example and researchers in construction management group use this
example very widely for explaining a simulation exercise. There is a simulation cyclic
operation here and for that they use this example very widely. So, what does this earth
moving operation example; let me explain this figure first. I am having an excavator, so
excavator is here, so if which does the earth cutting operation. Then the next step once
the earth is cut so, primarily I am just doing my earth excavation work and cutting the
soil, the first layers and then depositing the soil in somewhere else.
The next is loading operation. So, I need a loader and a dump truck for loading all the
excavated earth onto the truck. So, it does only the loading operation. So, after one truck
is fully loaded up, the next truck comes on the loader keeps on loading up onto the next
truck. And the next is hauling operation hauling is nothing but moving on road with the
truck. So, this is primarily the hauling operation, the dump truck is actually leaving the
site it is going to dump the soil somewhere. This entire process I have I have shown it on
a diagram right now. So, I have a project site. So, if the project site is the essential
resource for me in order to do the earth moving operation. So, I am having first activity
called earth cutting, then I have loading, hauling, dumping, return and then the trucks
return back, empty trucks return back to the loading site wait in the queue in order to
have the next loading operation to go on and the soil is dumped in a place in an marked
destination b.
Here what happens the movement of earth. Let me erase all this. What happens to the
movement of the earth? The earth movement is starting from here, it is moving on,
cutting, loading, hauling and then dumping. So, this is actually the earth movement
which you stopping till here. Excavator there is no moment here, but what happens to the
loader? The loaders moment is only till here (Refer Time: 29:22) only till here only till
the hauling operation and the truck it actually moves in a cycle. So, this actually shows
that truck cycle also. Loader loading operation, hauling with the moment of the truck
with the soil or the earth and then it dumps, then the empty truck comes only the empty
truck with no earth it just comes back and then stays here. So, there are now so many
activities which you have to match now. And three equipments are there in this exercise.
Now, let us start the example now. Now what is the duration? I am just starting up with
the simple example. What is the duration of excavating 3000 meter cube of earth work
using the following three equipments? I have now excavator, loader and I am using a 3
truck a combination to match up the loader also. This is the first cut decision we are
made and the quantity of excavation to be done is 3000 meter cube. Now what are the
production rates? Production rates can be obtained from the manufacturers also.
Production rates of excavator is 200 meter cube per day, of loader is 250 meter cube per
day, of the 3 truck combination is 150 meter cube per day. Now what will be my answer?
So, duration of the activity now, for the excavator alone if this is the production cycle
then it will take 15 days to finish the 3000 meter cube. Loader 12 days it will take to
finish the 3000 meter cube, 3 trucks combination it will take 20 days to finish my 3000
meter cube. And hence, what about the actual duration of this whole activity? So, it
needs 20 days, because a truck cycle also has to be matched up. So, I need 20 days to
finish up my entire exercise on this 3000 meter cube of excavation. Now what do you see
here, one just excavator is used for 15 days. My trucks are completely used in the 20
days duration, my excavator is used only for 50 days and 5 days its idle and what about
my loader? And the loader I am using only for the 12 days and the 8 days I am going to
pay for the idle equipment which is going to rest in my site.
Now let us rethink the combination. So, loader is underutilized which is what to has been
identified. If 4 trucks are planned to be used what will happen to the productivity? The
productivity just use interpolation. So, you will get at least 200 meter cubes per day and
the duration of this truck 4 trucks combination will be 15 days. So, what will be the
actual duration of the whole activity now? 15, 12 and this 3 trucks will be changed to 4
trucks and I will have 15 here. So, the duration of this entire activity will be now 15
days. So, loader is not utilized only for 3 days which is not too bad a option compare to
the earlier case.
Now what happens is there is a flow process here which is what I wanted to see here and
when I was explaining this example also, I showed all the flows. Why? Because when
you are matching, when you are deriving the duration of excavator, loader trucks and so
on, you cannot view everything as independent activity. You have to match with the
flow. So, that is where I showed the cycles I explained the flow process also. In the
earlier case also why I brought in scheduling estimating techniques is you cannot look at
one activity and see only that and then go ahead, you have to match the whole process.
So, for example, the flow process has to be matched.
Now I am giving you one critical situation right now. Let us assume the duration for the
haul dump return cycle it. Let us assume it takes 15 hours. Now what happens here, I am
having my excavator with just cuts earth and the loading is operation is going on. So, 12
days it takes to finish up the process and how many trucks I am going to use; four trucks.
So, when the first truck is gone, second truck comes and it goes on then what happens by
the time the fourth truck is supposed to be over, the first truck which has gone on for
dumping the soil has to come back and wait in the queue. Suppose if the duration for the
haul dump return; we are not very clear on what is the duration it takes. In peak hours it
may take 25 hours, in no peak hours it may take 15 hours, some situations like this arise.
What will happen now? You may have to make the loader wait and the excavator also to
wait and you may have to pay for other equipments. So, under those situations you may
have to think off increasing the loader and the truck combination or you may also have to
think of an option wherein the loading can also do the hauling exercise.
So, there are several options available. Accordingly, resource has to be changed,
modified and then you have to match up with the flow process. So, which is what I
would like to tell here right now. Now in the same way, Is the flow process, matching up
the flow process is it the only one way it works in all projects? No not necessarily, the
previous example it works and what happens in this case this was setting up a laboratory
example which I explained in the recently.
In the same example do you think the flow process will be matched up when you are
putting in more crews and more workers. It will not work out in every time. So, now,
what are the other options available now?
Let us again I am just giving a critical situation here and what we say is a milestone
planning. The resources all has to be based on the milestone which you are planned and
assigned in the project and accordingly you plan or you schedule the entire exercise. Let
us assume in the same developing a construction management laboratory exercise,
January 2020 the lab has to be ready for the classes and today we are in July. So, hardly
5 5 months or 6 months we have in setting up the laboratory.
Now what are all the critical paths one has to look out? We have to look from the
backward, from the finishing phase we have to see. So, first software procurement,
training, installation has to be done before the class has to start. To do the software
procurement, training and installation computer hardware purchase installation has to be
done and to set up the computer you need the table unit and assemblies, because you
cannot dump the computer on the floor. So you need the table unit has to be come in
place and before you put the table unit and then do the installations and so on networking
points that trickle points has to be fixed.
So, now, let us look at all the tasks. So, every task will have a list of sub tasks. For
example, electrical work I may have to do light points, wiring for the light points I may
have to do power points I may have to look at AC points, again earth work, wiring for
earth everything you have to do. Distribution box I have to set up. Ceiling lights I have to
fix. Ceiling fans I have to fix. Otherwise you cannot put the computer and then do the
installation also. Now if I assume approximately some n is a duration for each and every
item if by putting in more crew you will not be able to do because most of them are like
sequential activities and some only you can do it in parallel.
Electrical work can go parallel when you are doing the table unit assembly, because table
will not recover any electrical fixtures. So, some works can go parallel and some works
can go sequential, but the duration of these activities are like fixed and you cannot alter
and there are series of steps in each and every item like this. So, in this how do you do
the scheduling and how do you go with the whole exercises? We may have to fix the
milestones and accordingly work backwards on your schedules and then go ahead with
the schedules. That is how it has to be done.
Now we will see more on now relationships. In the flow process or in milestone planning
whatever it is now, we have learned little on durations. So, we will see now what is the
relationship that governs because your durations alone is not essential in order to fix your
schedule of the whole project. Now relationships, what are the different relationships
available? So, there are sequential and parallel which is a common relationship which we
use in construction phase alone. There are other types of relationships are also there that
we will see later and in resource relationships are also there which also we will see little
later.
Generally, when you see a dependency relationship by default it is call logic relationship.
So, what do you mean by logic? It is an order in which the activities are to be
accomplished. It is an order in which A has to be done first. After A, B has to be done or
A and B can be done in parallel. So, it determines the order in which the activities has to
be executed. And what is it exactly defined as a logical relationship which exists between
two activities. When the start, it can be also finish of one activity depends physically on
the finish or start of another activity.
Why am I underlining here physical, is this supplies only when you are using a logical
relationship. For other relationships this physical will not play a role the other resources
will play a role for resource relationship, information will play a role for information
driven relationships and so on.
Now there are different terminology which I would like to say. Number one links, links
is the word which we use when we are not following in the conventional finish - start.
For example, in the previous place I said finish of one activity it is connected with the
start of another activity; there is a finish there is a start. But in links what happens is I
can actually start another activity after some pace of the activity has been progressed. So,
it can start somewhere in the middle, it can also finish somewhere in the middle. They
are primarily called as a links.
Then the next loops are cycles when two or more activities are actually connected in a
circular manner, then they are called as loops or cycles which we will see more elaborate
as we move on into the week 2 week 4.
The next is redundancies; redundancies are unnecessary relationships in a project. I will

show you with an example also on this and the logical relationships should not be
confused with constraints. We will we will explain now what is a constraint.
Constraint is when an activity is subject to constraint such as approval of an owner or a

government agency etc. The constraints when you work on softwares like MS project or
Primavera you will come across some sort of constraints which we classify as flexible
constraints, moderate constraints and an inflexible constraints.
What does a flexible constraint? It is very flexible as soon as possible, the excavation
work should start as soon as possible there is no deadline. There is no demand. I can start
today if possible, I can start next week if possible there is no harm; so, if because it is a
flexible constraint.
Moderate constraints start no earlier then so which implies I can start the activity, but
there is a deadline only for the start, I can start even very ahead of whatever I can. Next
inflexible there is a date given and you have to go only with the same date, must start on
must finish on are all called inflexible constraints.
Now, moving on into glossary in CPM, which I will start with the small example. This is
a small example which I thought with that I will explain. Now, this is a simple example
and I have written ABCDEF, how to draw this network, how to represent (Refer Time:
41:41) this we will see in the next class, but this is a simple network diagram. We are
actually moving into network schedules on critical path method and before going into the
network schedules and critical path method some terminologies in critical path method,
then we will go with the proper introduction on that.

So, there are two terms in any network schedules; one is called activity and you know
what is an activity. Any activity is something which has a definite start definite end. So,
there is a time duration for this and there is a duration for all these activities then it is
called as an activity. An event is just a single step in the whole exercise. It can be start of
the activity or it can be end of the activity. That is primarily called as an event. And the
milestones are nothing but significant events. So, event will not consume any resource
and event will not take any duration. The same case with milestones also, they will not
consume any resource or they will not take any duration as such.
The next predecessor activity and immediately predecessor activity; I am having an

activity called G; I am just taking this example as G. So, for G,F,C,D all are predecessor
activities, but F is an immediately predecessor activity. The activity which just proceeds
the current activity, we call as IPA and all others are primarily call the PA’s.
Now, same case successor activities; I am just choosing an activity called D, F, G. Both
are successor activities to D, but F is the immediately succeeding activity to D.
Next redundancies for which I have to erase. Redundancies. how do you explain
redundancies? Redundancies are nothing, but unnecessary relationships. For example, if
I am marking from C to G, is it really necessary because already after C is completed,
you are starting and F activity. So, it is really not necessary if you are having a
relationship from C to G, unless it is a different type of a relationship. In logical
relationship the C to G is not necessary and this we call it as an redundant relationship.
The next is dangling; dangling is nothing, but the free lose activities which are in the
project. I just showed this example purposely. So, as such what should happen is all
activities should be connected. So, in the terminology A is supposed to be called as start
dangling and G is supposed to be called as finished dangling which you will learn more
when we are doing the PDM the precedence diagramming method.
The next, so what is a network diagram or a schedule? So, network diagram or a

schedule so, we are just coming into network diagrams. It is a logical representation of
the activities or events that comprise a project. So, primarily the way with which I have
drawn this complete exercise, this is primarily called network diagram or we can also
call network schedules when you are planning to write a duration here, 6, 5, 3, 4
something when you write then it becomes a network schedule also.
And what do you mean by path? Again, I will erase. Path is nothing, but the connection
between all activities. For example, here how many paths I have? I have A-B is one path,
CE is one path, C- F - G is one path, then D - F - G is another path. So, these are all
called the paths. Suppose AB takes 10 days to finish, CE takes 12 days to finish, this
takes 15 and this takes 14, then the CFG is called as a critical path, because it takes a
longer time to finish. That is called the critical path.
In all network diagrams you may draw the diagram and then when you are assigning the
duration to all these activities and then doing analysis, primarily the forward pass and
backward pass, then you primarily land up in arriving at or finding out which is a critical
path. So, that becomes a longest path in the whole project. Then crossover symbols,
suppose I am crossing over two relationships like this then there are symbols available
for the crossing over.
One is pipeline types and the other one is broken arrow which implies just like this. So,
these are the two forms of crossover symbols we have for crossing over. You can still
cross over two arrow marks. Otherwise if you want to use symbols the two symbols
which are available or pipeline types and the broken arrow types.
So, now so far what happened is in this particular lecture we have described all the inputs
in scheduling, starting from arriving at what is an activity, list of activities, relationships,
duration, duration little more we have seen in estimating techniques in a project as well
and we have started up introducing on critical path method, the glossary in critical path
method and the network schedules. The next class we will see more on critical path
method in detail and how to do analysis scheduling, how to draw a network and so on.
Thank you.
Lecture - 03
Critical Path Method
Welcome you all to lecture number 3, in the week 1 series; topic of today’s lecture is
Critical Path Method. So, in the previous two lectures I have primarily covered on the
general aspects of project management, construction management and so on and we have
also discussed on what are the basic inputs required for scheduling, and all the pros and
cons with all the basic inputs different methods available we have seen all those.
So, today’s lecture we will be covering on the first logic driven scheduling method,
which is very famously called as Critical Path Method or CPM.
So, in this class we are going to see the terminology used in CPM and the two forms of
CPM representation; one is Activity on Arrow diagram and the other one is Activity on
Node diagram. So, both the methods we are going to see in today’s class, along with
their scheduling presentation and the network analysis.
So, CPM network representation generally has two terminology: one is called nodes as
you can see here; the nodes are generally in circle or in rectangles and the other one is
called arrows. So, in the two types the way you represent the activities along the nodes or
arrows we have the two methods.
So, if you see here the types of network diagram, there are two broad categories: one is
called Activity on Arrow diagram very famously called AoA and the other one is
Activity on Node diagram and which is a AoN method. The other names for AoA are I-J
method Arrow Diagramming Method and in your AoN method we also called
Precedence Diagramming Method, but PDM is very generally referred for an advanced
version of an AoN. So, this is all on the background.
Now if you see the network representation, as I told you where you represent the
activities there in only you have the two methods. So in the first method activity on
arrow diagram. Now activities are represented along the arrows as how the name also
implies. So if you see here, I am representing the events along the nodes and I am
representing the activities along the arrows.
So, activity A is represented here and i is the tail event and j is the head event. So,
primarily any activity I do not represent an activity, I primarily use the events to talk
about the activity and that is a real reason why we also called the AoA method as an I-J
method.
In the AoN as the name implies activities are on the nodes. So, these are the nodes and
activity representations are generally done on the nodes. So, here if you see; so this is the
nodes and activities are represented here and the arrow marks generally show the
relationships between the two activities. So that is how we have the two methods and
these are the two representations for the same.
Now, little more on CPM network schedule. So, let us look a little on the history. CPM
network is an very old method almost like 1950s or so it started and still going on. The
first method which came in existence was a AoA method, because of little problems with
the dummies and etc, so AoN came into existence and people are very comfortable in
using both the methods as their choice.
Now what are the assumptions in drawing the CPM networks? There are few
assumptions: the arrows whether it is AoA or AoN are generally drawn from left to right.
And the orientation either horizontal or vertical there is no backward links or something
and no curve lines are also available. And there are fewer formats for network
representation.
For example, arc format which is not that predominantly used in practice. The other one
is called sloping format. Sloping format generally has horizontal lines and also sloping
lines which we call it as an angular lines for representing the activities. So, it is
represented like this. In staff format the same thing what we do is we represent in terms
of horizontal and vertical lines only. There are no sloping lines and as such there are two
ways of presenting your diagrams.
The length of the arrow generally has no significance on duration. Apart from methods
like timescale diagram, bar charts and so on, but they are not called CPM network
schedules. Generally, any network schedule the length of the arrow is insignificant of the
duration.
Now Activity on Arrow diagram.
So, we will see little more on that. So, the guidelines in drawing. So, primarily you have
head events and tail events, as I tool i is primarily the tail event and j is the head event.
Numbering is generally done from left to right which implies it just starts from i and
proceeds to j-th event and so on. And the length of the arrow as I have told already its
there is no significance on duration, even though you are representing the duration along
the arrows.
Then there is something called dummy activity and as I said it is a drawback in AoA and
for which only you have an AoN diagram into existence. So, what is this dummy
activity; dummy activity is generally introduced in a CPM network either to show a
dependency relationships or to avoid duplicate numbering of activities. For any activity I
should have an unique i and a j and this i-j cannot be shared for another activity, so for
that reason. And the other reason is just to show the logical dependency between the
relationships there may be a requirement of usage on dummies. These are the two
reasons in the examples which I am covering I will show both the examples, and how to
use for and where are the dummies coming in and what is the purpose of the same.
Dummies are often shown by dotted lines. It is easy to identify. It consumes no time or
resources.
There is something called Fulkerson’s rule, so that is talks about the forward numbering.
So, i always has a number which is less than j and there is also something called skip
numbering adapted for large complex projects. Every project whether it is medium or
large project generally where changes are unavoidable and if there is a change of scope
of the work and few activities are added up what happens is, you may have to redo the
numbering if you are planning in consecutive 1, 2, 3, 4, 5, 6 and so on.
So, general advice is to leave some numbers or gaps in the middle and start like 10, 20,
30 or 10, 15, 30, 50 and so on, so that you can add activities with the change in scope of
the project. That is the main reason on skip numbering.
So, these are the general issues with AoA diagram.
Now, let us take an example in order to really understand how to represent a AoA
diagram and the analysis later on. So, first example primarily is a sewer house
connection for collecting sewer from households and how it links to the public or streets
sewer.
So, if you see here this is a schematic diagram, I have a main pumping station and there
is a sewer treatment plant; STP stands for Sewer Treatment Plant and MPS is for the
Main Pumping Station. I have interceptors as shown here and there are lifting stations
also, there are pumps for pumping up the sewer collected also and there are branches
from the house it just moves on and so on. And inspection pits from the pits connections
are all happening. Now.
So, a little more on the layout. So, there are pressure flows, there are gravity flows in the
area as well. So, you will have location of manholes like this, lifting stations are all here
located at station 1, station 2, station 3, and there are manholes also based on the slope
and gradients and the pressure on the flow of the sewer.
You may have to do either gravity flow or there can be pressure flow and accordingly
there are lines marked here on green and red and there are several existing lifting stations
also, new stations are also planned for connected with the with the existing stations in
order to do the pumping action and collecting of the sewers to a line, right.
Now the scope of the present project is only on DWC pipes; double wall corrugated
pipes just for one manhole I have and the pipeline which I have as a connection. So, what
are the activities, how was the duration calculated on, how was the construction really
done?
So, let us talk only about the underground pipeline including the manholes. So, how do
you plan for the whole section of the work? So, scope of the work is planned is 240
meter sections. So, covering the entire flows I have 240 meter of a stretch I am planning
to take, so how much time it generally takes for finishing of the entire tasks? Working
hours per days by default is 8 hours per day which is also taken. Manholes general norms
in this particular project was to be provided at 30 meters interval. So, approximately
there is 8, starting with the connection or the starting points you may have to do 9
manholes in the whole sections, ok.
So, now the activities for this particular construction, there were six activities we have
taken it was not completely exclusive, but we just took out only the major activities for
the whole project. So, excavation, dressing and ramming, brickwork for the manholes,
pipe laying and joints, then backfilling and compaction and the surface PCC; these were
the 6 activities that were taken.
In order to have an easy way of writing IPA’s and going ahead with the analysis we also
introduced an activity ID, so A B C D E F. IPA I have assigned based on the activity IDs
of these activities only. If you see here and all activities are done in sequence except the
activities B and C which have shared the same IPA, so they can be done in parallel ok.
So now, what is the crew productivity? So, this is not rate per crew this is a crew
productivity, the productivity of the following crew that was planned for the project was;
for example, excavation 30 meters per hour and the crew size available was 1 crew size
and the crews available were 1 crew; which implies one group of only one worker was
available for the entire excavation activity.
So, duration taken is a simple formula which you always know; what is the quantity of
work to be done? Which is 240 meters divided by the productivity constant which is 30,
so this will be like 8 hours is the duration for excavation. Now for the second activity
dressing and ramming. So, 20 meters per hour and there were two workers were
available in one group and one group was only available at that time. So, one group of
two workers, so it is primarily 240 by 20 which is actually equal to 12 hours is a duration
calculation.
So, similarly all the durations for all the activities were calculated in that way. This is
only to show how to represent, how to define the scope of the work, how to list out the
activities and how to get the duration. In this case the resources were primarily driving
out the duration and the duration was not fixed as a constant and then the resources were
are same. In this case the duration came out naturally as a result of the crew size which
were available, ok.

Now, this project is represented here. So, I have the list, so from the previous table I have
taken up. I need only these information, so I am just showing up only the IPAs and the
duration for all these activities, ok. As I told the activity B and C are in parallel.
So, I have not shown that activity, but I have shown all the other sequential activities in
the network. And the duration and the activities are represented along the arrow marks.
Generally, activities represented above the arrow mark as shown here and duration is
generally written below the arrow mark as seen here and numbering is also done.
Now, in this particular case. So, this is primarily I am going to introduce a dummy,
because if I connect the two activities like this what will happen is; I am going to share
the same i j for both the activities. So, I wanted to avoid that. So, what I have to do is
there are several options for; there are several options for showing the dummies
primarily I have to introduce the dummy.
So, one option is I can introduce a dummy here and then I can connect the activity, this is
one way. The second way is, so from 2 to 4, I can introduce the activity and then I can
connect the dummy. This is activity B, this is activity C, this is activity B, this is activity
C.
Other option is; so two options are already shown here.

The next option is I am going to connect my activity B as such. So, this is my activity B.
Introduce a dummy here and then connect the activity which is C or else I am going to
introduce my activity here and then connect my dummy, ok. In this way how many
options I can have 4 options of putting the dummies and all 4 options are equally good. It
is primarily based on the convenience with which you are going to choose the options.
So now, what I am going to do is, I am just going to choose one of them here.
So, this is primarily the option which I have chosen and I have just represented the
network, but there are other options also available and this dummy is introduced
primarily to avoid the unique numbering of activities between the activity B and activity
C. So, that is all on the AoA diagram.
Now let us move on to the other form of CPM representation which is a AoN diagram.
So, the same example I am going to take. So, the same example I have taken: duration
everything I have taken from the previous example which I was listing out right now.
So now, how do I represent this example; as I have told earlier, activities are generally
written above the inside the circle but above, and the durations of those activities are
written below, but inside the same circle. And generally, we use circle representation for
a for a CPM network and we use a rectangle representation for a PDM network, but the
choice is among the users you can use anything.
So, this is how the activities are represented. So, after A, I have B and C in parallel and
after B and C, I have to start my D. After D is completed E will start, and after E is done
F starts and finishes.
This is a whole network. As you can see here this is really an easy network to draw
compared to a AoA because of the absence of dummy activities, ok. So, that is where
this is most preferred compared to the AoA method.
Now moving onto network analysis. So far we have seen network schedule, the
terminology when you say network schedule is primarily representing your networks on
to the with either in the AoA format or in the AoN format and the minute you say
analysis, primarily you are going to do forward pass, backward pass, and critically you
are going to find the critical activities in the path. So, network analysis is computation of
forward pass and backward pass and identify the critical path.
Now, why we do this calculation and who wants the float or the slack; now there are two
terminologies which are coming in: one is called float other one is called slack. Float is a
word which we generally use in the AoN format and slack is a word which we generally
use in a AoA format, after few slides you will understand that. Primarily what is it? It
tells you how much is the delay in time, how much you can actually delay any activity
without affecting the final project duration of the whole project ok.
Now, you know that there is a buffer time existing for a few paths or for a few activities
and so on. Now who has the responsibility or who owns that free paths in the luxury of
starting and activity or a particular stretch, late or delaying and slowly progressing and so
on. Is it owner or contractor? It is primarily definitely not the owner and definitely not
the contractor, it should be a project related float, ok. So, that should be a good. if you If
there should be a successful projects to happen then the project should own the float and
neither the owner and not the contractor.
Now, let us see the first method which is a AoA method and the forward pass, backward
pass, and the slack calculation. So, forward pass it is a computation method to find the
early event of early event time of events, that what is the EET; Early Event Time. And
backward pass it is a computation method to find the late event time of event, so which is
LET and slack is nothing but the spare time in the entire stretch of sequence of events,
and generally slack is given as LET minus EET, ok.
Now, how do we represent? For representation AoA, this has been discussed earlier. For
analysis how do we represent? This numbering is not generally used when we are doing
analysis. So, the late event time or early event time is generally represented inside the
events and then slack actually comes on one side, ok. So now, this can be also altered.
So, LET and EET.
So, inside the circle I can also have EET and LET, and I can also write my slacks this is
also possible and both the methods are equally fine, ok. So, whatever you want you can
choose here, but generally this is a way to represent the analysis.
Now, let us move on I am just going to take a hypothetical example right now in order to
show the analysis in detail, ok. So, I am just having a few activities here represented
alphabetically A, B, C, D, E, F, G, H, I. And the immediately preceding activities (IPA)
for these activities are given as like this ok. And I can also have table which can also
equally have an ISA as well ok.
So, it is primarily again based on convenience, but the default relationship procedure is
to show it on IPAs only. And then I have duration for all these activities given here, ok.
Now what do I wanted to do? First step last example was very simple so it was easy for
you to just represent the activities. This example it is little big, as and when you move on
in take a real construction project you may have 50 activities are even more. So, how do
you start and represent the activities? This is especially good for beginners that is
something called sequence steps.
So, first step is to identify the sequence step for the entire table. So, that is easy for you
to represent the activities and then connect the activities so that you have a neat diagram
in a very few round, ok. So, sequence steps how do you do? So, those activities which
does not have an IPA we will come in the sequence step number 1. For example, A, C,
and D that is not have an immediately preceding activity.
So, they all will form the first activity to start with, ok. All those activities which had SS
1 we will all be become as I predecessors we will have the sequence steps number 2. For
example, A, C, D had SS 1s, so those activities where all its coming as a predecessor. So,
primarily activity B, activity E, activity F we will all will form as a sequence steps
number 2, ok.
Now, what all the activities which have sequence steps number 2 as an IPA will all
become the sequence steps number 3. For example, so it should be E F or B. So,
primarily the activity G and H is from is under sequence step number 3 and G and H is
IPA for I and hence that becomes a sequence steps number 4. In this way if you progress
it is easy for you to finish up the entire arrangement of the activities.
Now what do you do after arrangement? So, first is AoA diagram.
So, the same table I have shown it on the top for you to really correlate and understand.
So, primarily the first sequence step, second sequence step, third sequence step and
fourth sequence steps are given here. So, I have arranged my activities accordingly. This
is activity on arrow diagram.
So, I am representing all the activities along the arrows. So, A, C, D is written here B, E,
F is the second sequence step; G, H is the third sequence step; and I is coming in the
fourth sequence step, ok.
Next is connect all these activities using a dummy relationship ok. So, seeing these
relationship which I have along the IPAs. So, A has no IPA and B has A, C, D as an
IPAs, so I am having A, C and D as IPA. Then C does not have any IPA, D does not
have any IPA. For E activity I have C and D as my IPA, for F it is activity D is my IPA.
Then for my activity G, F alone is my IPA; for activity H, I have B, E and F and then I
have for activity I, G and H are my predecessors.
So, connect all these with the help a dummies; that will be the second step which you
have to do, ok.
So, the diagram is something like this. Now once you have connected all the
relationships, then see whether you can remove some of the dummies so that your
relationship is still maintained, but unnecessarily you are putting all the dummies can be
really avoided ok.
So, in this particular case for activity E, C and D are the predecessors. For activity B; A,
C and D are the predecessors. So, keeping only these two dummies in order to represent
the logical flow. The first example I showed a dummy representation to primarily
showcase the unique representation of activities. In this example the dummies are used
for showing the logical relationship between the activities.
So, the next step will be like; now check for these relationships. Again these 3 dummies
can be easily removed, so I have removed all the dummies. And to maintain the
relationship, so B, E and F is primarily for H and for F it is for G and then I have I which
is H and G. So, when you are removing the dummies see to that and correct all your
IPAs whether it is right or wrong and then confirm for the figure before you are
removing it.
Now check for each and every activity, I think all my activities are checked for all the
dummies. So, this looks like a final network, ok.
Once the final network is arrived at then the numbering generally happens. Numbering
here I have given in sequence only. So, 1, 2, 3, then 4, and then after 5, 6 because I
should be always less than G and then i have 7 and 8. So, I have a done my numbering
also, ok. And now paths: I have so many paths in this entire network, ok. For example,
there are seven paths in this network A, B, H, I C; so this is primarily A, B, H, I ; then I
have C a dummy B, H, I. Then I have C, E, H, I; then I have D, B, H, I; D, E, H, I; D, F,
H, I; and I have D, F, G, I also. So, these are the different paths I have in this particular
network, ok.
So, path explanation also I am just showing it along with this case. Now the main issue
behind us now which is a critical path for me and what is the duration for all these paths,
ok.
So, now let us start the analysis segments. So far I showed you the simple steps in order
to arrive at representing a network, now we will start with the analysis. The same
diagram which I have taken up earlier. As I told you numbering will not be done when
you are doing the analysis part ok.
So, I am just taking the example and as is shown you earlier the representation for
analysis I have divided the circle into 3 segments: one for EET, one for LET, and one for
the slack. Forward pass is primarily to calculate the early event time of the events ok.
So now, how do we start? As usual, so you have to start the first event at 0, and again
this also I forgot to say earlier. So, these three, and every time you should see to that
there is a one starting event and one end event as here, instead of having multiple stars
and multiple lanes to avoid dangling of activities. So now, here I am starting with 0 ok.
So, D has a duration of 6.
So, for this particular event the early event time is 6 here and for activity C; so 0 plus 3 is
3, this is 6 plus 0 is 0, so maximum is 6. So, I am going to keep this as 6. Now coming
back to A, so this is 0 plus 6 is 6, this is 6 plus 0 is 6. Again there is no minimum
maximum, but generally you are supposed to take maximum.
So, this becomes 6. Now for this particular activity I have 6 plus 3; sorry I have to do this
because this is in the successor list. So, 6 plus 4 is 10, so this early event time of this
event is 10.
Now, coming here I have three arrows connecting to this event 6 plus 3 is 9, 6 plus 2 is 8.
As a result of 6 plus 3 this is 9, 6 plus 2 this is 8. And this is 10, 10 plus 0 is 10 here. So,
I have 3 event numberings here, maximum only you should take when you are doing the
forward pass, so this becomes 10. Now if you look at here 10 plus 6 is 16, 10 plus 3 is 13
so maximum is 16 so this becomes 16, and 16 plus 5 is 21. This is primarily the forward
pass.
Forward pass what you do? Early start for the first event generally we take 0 and for all
other events is maximum of the preceding events plus the duration, ok. So, that is what
we have to take here. Next is backward pass: in the backboard pass we have to generally
start from the last late event and then we have to proceed backwards and late evens are
generally for the successor activities is primarily minus you have to subtract the duration
of the activity and go ahead. So, backward pass is generally done in this way and your
forward passes are generally done from left to right.
Now we are seeing the backward pass so this is 21, 21 minus 5 is 16 here. Now there are
two lines which are two preceding activities after this event. So, 16 minus 6 is 10 here
and this I have on more arrow so I cannot do this, so 16 minus 6 is 10.

And if you look at here I have 10 minus 0 is 10 and 16 minus 3 is 13. And when you are
doing the backward pass we have to take the minimum of the two numbers. So, this will
be 10 here I have written 10 here. So, 10 minus 3 is 7 here, then 7 minus 0 is again 7, 7
minus 0 is again 6 here. So, this will be 6; now 6 minus 6 to 0 here 7 minus 3 is 4, 7
minus 6 is 1. So, I have 3 event times and the minimum I should take which is 0 ok. So,
that is how the backward passes are done.
So, forward pass is generally calculation of early event times taking into the activities
which are all preceding, and you should take the maximum of the value and then go
ahead until completion. When you are doing the backward pass you have to start from
the last event and then proceed till the first event and every time you have to subtract
from the activities and go backwards with the arrows. So, you have to start from the head
event and then proceed to the tail event every time, ok. That is primarily the backward
pass. So, I have got LETs and EETs.
Now, the next procedure is to find the slack. As I told you the slack formula is LET
minus EET ok. So, 0 minus 0 is 0 here, 7 minus 6 is 1, 7 minus 6 is 1, 6 minus 6 is 0, 10
minus 10 is 0, 10 minus 10 is 0. So, you have to show for all the events, ok. Once you
have done all these calculations primarily the evens which have 0 as a slack are all on the
critical path ok. So, primarily what happens; this event is having a 0 slack, this event is
having a 0 slack, this is having a 0 slack, this is also having a 0 slack this and this, so all
these are having a 0 slacks, ok.
Now if you see the path, so how many paths I am able to build? I can build D, F, H, I ok,
on the other path is D, F, G, I ok. Blindly if you go only with the slack values and then
go ahead and fixing up your critical path I may land up in the assessment that there are
two critical paths in the entire network, ok. But is the logic really true. If you see here 10
plus 3 is 13, 10 plus 6 is 16, you really know that G is not on the critical path, ok.
Now coming back to the drawbacks in AoA method, one of the drawback we have
already discussed what is it; the presence of so many dummies in the network only to do
unique representation of activities or to represent the logical relationships. In this
network I had 3 dummies, in the last example I had 1 dummy. So, dummies are one
biggest disadvantage in a AoA diagram.
The other disadvantage is: sometimes you may be misguided with the values on the slack
calculations, ok. What to do in these cases? You may have to again work out on all these
activities and cross check for your values whether they are on the critical path or not and
then you have to go ahead with the calculations. This is another biggest disadvantage in a
AoA network, ok. That is where people started preferring a AoN method.
Now let us show the AoA representation again. So now, the critical path as I told you G
is not on the critical path, so the critical path is D, F, H and I along with the dummy in
the middle and the duration for this particular project has taken 21 units of time. That is
how the critical path calculations are done, ok.
Now in the AoA method whether you are using or the AoN method the procedure for
computation on analysis is same. First you have to do forward pass next backward pass
and the flow calculations, in AoA we have slack calculations ok. That you should keep in
mind.
So now forward pass: it is a computation method to find the early start and early finish of
activities. There the focus was events, we were finding out early event time and late
event time of events in AoA diagram. Now we are going to see early start and early
finish of activities, this is forward pass. Backward pass same way it is a computation to
find out the late start and late finish of activities ok. Short forms are ES, EF, LS and LF.
In AoN as I have told you we generally have terminology called floats to show the free
time or the delay with which we can have a delay the start of the activities ok. So,
primarily the two very famous floats we generally use to find critical paths are total float
and the other one is free float. Total float is also called as a path float and free float we
also call it as an activity float, you will see when I am explaining with an example ok.
Total float it is a maximum amount of time and activity can be delayed from its early
start without delaying the entire project. So, this actually talks about the one stretch of an
activity or one particular path, how much is the buffer or delay an activity can do without
delaying that particular path in order to finish the in order to without delaying the entire
project duration. So, the formula for calculation is either late start minus the early start or
late finish minus the early finish of the particular activity.
Free float as I told you it is also called activity float. It is defined as a maximum amount
of time an activity can be delayed without delaying the early start of the successor
activities ok. So, the formula is minimum off early start i plus 1 which is the next activity
minus the early finish of the current activity. That is the formula we generally use here.
In the free float what happens is only tells you what is a buffer time or delay I can have
on the activity without effecting the successor activity on the whole network. That is
what is the free float meaning behind, ok.
Now, let us see the representation. So, representation for AoN is activity on the die on
the node diagram, this is known to you already. Now for analysis how do you represent?
So, primarily we have two lines close to the activities. Why these two lines are only to
show that that is for the particular activity; if you have a very clumsy network very big
network and you keep putting all the 4 values then you will not know for which activity
these values are represented for; after the entire calculations are done ok.
So, the early start of the particular activity A, then early finish, this is late start late finish
and early start early finish, similarly for B. And total flow generally rectangular blocks
on the top of the activity and free float is given by a small triangular below the activity.
Again, these two symbols are only to avoid confusions and to show this is for the total
float and this is for the free float. When I am having an activity which is below and I am
writing two values, then with the symbol I will know that these are the total float on the
other activity and this is free float of this particular activity. Only for that these
representations are given.
Now, with this we will go back to the same example, not the same I would say this is an
other hypothetical example. So, sequence steps as earlier I have given. So, these are the 4
sequence steps and all these activities are listed as per the sequence step as per this
figure, I am not expanding the sequence steps again because this is known to you ok.
Now relationships, next step is to connect all the relationships for all these activities.
Here I have moved the activity G down, because I know it will be crossing if I am
putting it on the other side so I have purposely put the activity G down, ok.
And this, since I have 3 activities in the starting I wanted to avoid the dangling. So, I
have introduced a project start, that can be in the same shape as the activity like this and
you can introduce as projects start ok; and that has a duration as 0. So, only to avoid the
dangling of the activities and this is my entire final network.
Now, first forward pass, as I told you earlier so primarily you have to put two lines close
to each other and forward pass is primarily again from left to right, calculation of early
start and early finish of all the activities ok. Project starts, so starting at 0 and the activity
duration is also 0. So, 0s this is also 0, so you generally have 0 ok. So, this 0 0 and 0 all
three will have the same 0 values ok. 0 plus 6 is 6, 0 plus 3 is 3, 0 plus 6 is 6.
Now if you see activity B there are two relationships before this activity A and C this is 6
this is 3 maximum time only you should take in case of forward pass. So, this is 3
comma 6 and the maximum is 6 here. Now for this activity again I have C and D as my
predecessor and the early finish of C is 3 early finish of D is 6, maximum only I am
going to take which is 6 so I am keeping here as 6. And this will be the same number
because there is no other relationships ok. Now 6 plus 3 is 9 here, 6 plus 2 is 8 here, 6
plus 4 is 10 here.
Now coming to activity H, I have 3 predecessors one is 9, one is 8, and one is 10 as the
result of B E and H F predecessors. The maximum is what I am going to take for the
forward pass, so this is 10. So, primarily I have to finish B, E, and F to start my H. There
is a logic to that also ok. If I start at 9 then F is not going to complete, so I have to finish
for all the 3 activities.
When will all the 3 activities get over that will be on day 10 only. So, I can start H at the
earliest starting time of H is day 10 because I want to complete B, E and F before that,
ok. Now this becomes 10 ten plus 6 is 16, this is same as 10, 10 plus 3 is 13. Here again I
have two relationships 16 and 13. Maximum is 16 so my value is 16 here ok; 16 plus 5
becomes 21. So, this finishes my forward pass calculations.
Now, coming back to the backward pass, backward pass always progresses from right to
left from the last event to the first event ok. In the forward pass I finished my early finish
at 21, the same value is generally taken for the backward pass late finish ok. The same
thing happens in the AoA also. The EET of the last event will be the same as LET of the
last event ok.
So, this 21 is here 21 minus 5 is 16, so this become 16, this also become 16. 16 minus 3
is 13, 16 minus 6 is 10 here ok. Now coming here so this is 10, this is 10, and this is 10.
And for this particular activity I have two values 10 and 13 ok. As I told you earlier I
have to choose a minimum when we are working on the backward pass.
What is a meaning behind this? So, this stops on, this I can start even on 13, but what
will happen the latest time with which I can finish should be 10 ok. So, that is the reason
why I have to finish the activity F latest by 10th day. So, all these are primarily 10 as a
result of this relationship; 10 minus 3 is 7 here, 10 minus 2 is 8 here, 10 minus 4 to 6
here, now this becomes same 7 so this relationship is here.
Now here I have 7, 8 two values are there minimum is 7, so I am going to take 7. Here I
am having 8, 6 same way I have two values, and the minimum is 6. I am going to take 6.
So, 6 minus 6 is 0, 7 minus 3 is 4, 7 minus 6 is 1. Now here 1, 4 and 0. Minimum value
0, so I am going to do this as 0 ok. So, this finishes my backward pass.
Now if you see here. So, once I have done forward pass. I have done my backward pass,
I know the early start, early finish, late start, late finish, of all the activities I have
calculated ok. The next step logically is to find the critical path.
To find the critical path we should calculate the float ok. Generally, you can do free float
total float. Both the floats you can calculate. But default calculation is by using total float
only people find the critical path. Wherever the total float is 0 in an activity those
activities will come on the critical path ok. Now I am having the same network. So, how
to find the total floats is primarily the late start minus early start or late finish minus early
finish is what I said. So, it is primarily the late start minus the early start 1 minus 0 or 7
minus 6. So, both are same here, total float I am getting as 1.
For activity B it is either 7 minus 6 or 10 minus 9 both you can use any one of them, so it
is 1. Here 4 minus 0 is 4, 0 minus 0 is 0, 8 minus 6 is 2, 6 minus 6 is 0, 10 minus 10 is 0,
13 minus 10 is 3, 16 minus 16 is 0. So, I have finished my total float calculations
completely.
If you look at this network right now where are the total floats 0 values. So, I have at
activity I, I have at activity H, I have at activity F and also on D. this logically will be 0
because 0 minus 0 is 0, ok. Now this is also one of the other way to check whether your
calculations are right or wrong. Obviously, in any network there will be minimum one
critical path in a network and the path in the sense it should be connecting the whole
project and there will be at least one path which is on the critical path ok. If it is not
coming then obviously we have done some mistake somewhere in the calculations, ok.
Now let us move on to representing the critical path. So, D, E, F, H, I, E is on the critical
path. Now let us go back to the earlier AoA diagram as well.
You also had the same network which is D, E, F, H, I, E was on the critical path. So,
generally people do not do both the methods for calculations, but I am just showcasing it
to you that you will get the same results whatever the method you choose.

So, the next is free float. So, how do you calculate the free float? Free float for any
activity is; so primarily the maximum of the early start of (Refer Time: 44:42) activity
minus the early finish of the activity ok. For this particular activity I have only one
successor and the early start is 6 minus the early finish is 6, so 6 minus 6 become 0 here.
For this particular activity I have two successors and both the values happen to be the
same, so 6 minus 3 is 3 for me.
For the same activity D, I have two successors 6 and 6 are the early starts of both the
activities and for the current activity I have 6 as my early finish, so this becomes 0 is my
free float, ok. Now for activity B again I have only one successor, so this is 10, 10 minus
9 is 1. For activity E, so this is also done. For activity E, I have again only one successor,
so this is 10 minus 8 is 2 here so this is my free float.
For activity F, I have two successors. The early start is primarily 10, so this is 10 minus
10 so this become 0, ok. Now when I look at activity H, so the early start on the
successor activity 16 minus the finish activity is 16, so 16 minus 16 is 0; so, this become
0.
Now for activity G, the early start of successor activity is 16, 16 minus 13 is 3 so this
becomes 3 for me ok and for this particular activity I, what you have to do is I have to
assume that there is one more activity which is primarily a finish and what is the early
start and the early finish if this activity exist. So, this will be 21 only, this is 21. So, if
you have to take that into consideration so 21 minus 21 becomes 0 here, so this become 0
ok. That is how the free floats are calculated.
If you look at the forward pass total float and free float, so whatever activity I have for
example, are on the critical path the same activities will be on the critical path only for
the free float also. So, you will have 0 for activity D, you will have 0 for activity F, you
will have 0 for activity H, you will have 0 for activity I. So even with free float also you
can still find the critical path, ok.
Now what is the meaning of the free float; that I will explain here. Let us as assume I am
taking an activity B, this means I have a 3 day duration for activity B this free float 1
means I can delay the activity B to 1 day and nothing will happen to the delay in the
critical path. That is a meaning of the free float I have here, ok.
But if you take in the total float values, the value of this total float which is again 1 this
means this value of 1, I can delay my activity B and because of the delay in activity B to
1 day it will not affect my A, B, H, I path to come to delay in the completion. That is
why it is called as a path float and in this case it is called as a activity float ok. So, that is
all on about the total float and free float.
So, if I wrap up on the AoA diagram and the AoN diagram; AoA diagram came in
existence first and then AoN came in. Drawbacks on the AoA I have covered up then and
there, primarily the dummies and also sometimes you know miss guidance with the
values on the slack values because we are calculating it on events ok. And so, in AoA we
have forward pass, backward pass, primarily with the early event time, late event time,
and the slack calculations.
In AoN diagram primarily activities are on the nodes. And we have the early start early
finish in the forward pass, late start and late finish in the backward pass. And two types
of float we have primarily seen one is total float and the other one is free float. And these
two floats are predominantly used in the AoN analysis.
So, that is all about the analysis and the scheduling method in the CPM network. Now
what are the other types of floats in the CPM network?
The next one is interfering float. These are all not much used, but there are other types
also existing. So, it is defined as the maximum amount of time an activity can be delayed
without delaying the entire project, but causing the delay to succeeding activities. So,
this interfering float is generally TF minus FF, which implies total float minus free float,
ok.
Next is independent float. It is a maximum amount of time and activity can be delayed
without delaying the early start of the successor activities all put together, and without
being affected by the allowable delay of all the proceeding activities, ok.
So, as the explanation goes the formula also is saying that minimum of all my early start
of the successor activities minus maximum of the late finish of all my preceding
activities. There is a mistake here. This should be LF ok. Of all my preceding activities
LFs and when I have all my successor activities the early starts, so minimum of the early
starts successors minus maximum of late finish predecessors minus duration of the
current activity will give my independent float, ok.
Negative float, sometimes what happens in the contract completion date, there may be a
definite date called finish on or something there is a definite constraint given there. And
sometimes the project expands or delays beyond that when you are putting all your
constraints, your resources, schedules and so on.
What happens your project schedule may go on a date which is beyond even the as-
planned schedule completion date. Then what happens in that sense you may have a
negative load coming into the whole project. That is primarily called negative float.
Then next type of floats is terminal float. So, terminal float is nothing but I have a whole
project and there is really a different time for my planned completion and also for the
contract completion, ok. In the earlier case I said if the contract completion or schedule
pressure completion is little early you start getting a negatives on the floats, ok.
So, let us assume I am having a project which is planned to be completed in 30 days as

per my TF calculations, forward pass, backward pass. I have done. I have all finished
everything, total positive duration for the entire project is coming out as 30. If the
contract completion time is said something as 25 months then what will happen is this 5
months primarily you will get us as a negatives float all along the whole activities.
Suppose if the project completion contract completion time is something around 40

months or something. So, what happens? I can easily finish my entire project in 30
months and the 10 months is primarily a real float which I have on the whole project.
This is primarily called the project free float; terminal floats are also called the project
free floats.
The next one is internal float. Every activity we are calculating a duration, based on the
resources, productivity calculations, crew size available. I am primarily calculating
duration for each and every activity. Sometimes what happens is that let us assume I am
planning for an 8 day duration for an activity ok. All the 8 days you need not be working
on the activity there may be buffer here and there may be a free time you can still take
out from the whole activity which is what is primarily called the internal floats. So,
within an activity itself it is possible that there is some kind of a float coming out that is
primarily called as an internal float.
The next one is contingency or buffer. So, contingency should not be confused with
floats. Contingency is purposely you are giving an allowance onto the project and that is
primarily called as a contingency. So, contingency or buffer is an allowance specifically
added to a schedule or a network to take into account of unforeseen circumstances.
So, it is an unusual event that may or may not occur. That is primarily you cover it up as
and contingency and cannot be assumed for ideal conditions when preparing the
schedule.
PERT allows contingency because it does contain 3 types of an values like optimistic,
pessimistic and most likely.
So, the network itself has uncertainty in the duration calculations and PERT is another
network technique which I will cover later. That does allow contingency in some way.
And most of the contingencies may have little or no impact on the project duration.
So, that is all on the lecture for today. So, today’s class we have seen critical path
method. The two types in the critical path method one is the AoA diagram, AoN diagram
the representation of these two networks., the analysis of these two in terms of forward
pass, backward pass, slack calculations in AoA, float calculations in AoN and the other
types of float also we have seen ok.
So, that is all for today’s class and, we will continue on precedence diagramming method
in the next class.
Bye.
Lecture – 04
Precedence Diagramming Method
So, the course outline or primarily the lecture outline for today’s class will be like this.
I will introduce, what is PDM, how to draw a PDM and so on. And then for network
analysis, we primarily have two different methods; one is contiguous activity execution
and the other one is interruptible activity execution. In the sense, when an activity starts;
in the first case it does not stop in the middle and it executes till the end, in the second
case it can have a break in the middle of the execution as well. So, we will see how the
analysis goes.
And now, what is PDM? So like CPM, PDM is widely used in construction industry. We
have seen earlier class, what is CPM; it is critical path method and this is also very
widely used and it uses only the AoN notation, which is nothing but Activity on Node
formation only. Now, what is the specific difference between CPM and PDM is, the
presence of four logical relationships and the lead and lag, you will see in the next slide.
So, what is the four logical relationships? Number one is finish to start relationship. As
you can see here, this is the most widely used relationship in CPM network, CPM, PERT
and so on.
So, this is nothing, but when an activity is executed completely then the next activity
starts, that is the meaning of after finishing the other activity is starting. So, primarily it
defines the relationship between any two activities. And how it is represented in a PDM?
As I told you earlier, the nodes can be circles or rectangles and generally we use circles
for CPM and rectangles for PDM, ok. So, I have represented nodes and as I told this is an
AoN representation, so activity on nodes. So, I can use this format or I can use this
format, so both are fine, ok. Then, the next relationship is start-start relationship and
finish-finish. Generally the start-start and finish-finish are used together to explain the
partial concurrency of activities, ok. This is start-start, after A starts, B should start ok;
that is what I am representing here. You can also use this, you can also represent like this
ok.
So, any form is fine and finish-finish; so this is a finish-finish. After A is completely
done, B should be completed and the same thing I have represented when it is arranged
in the vertical order. Next is start-finish relationship, which is these two figures. So, this
explains after, when activity A start B should finish which is very rarely used in
networks and it is not that widely used ok. So, these are the four relationships we have in
PDM networks, ok.
Now as I have explained you here, so always the start part of an activity is only this and
the finish is only this. So, if I want to draw a network ok, I can use my arrows anyway
from here ok, it can end up here. So, this is primarily the start portion and when I want to
end it, this primarily called the finish portion. So, I can do like this or I can start from
here like this, ok. And never you should use this segment of the whole box, anyway the
meaning of that you will understand later.
So, next is lead and lag. As I told you the two main functionalities in PDM is, the other
types of relationships and the lead and lag, which is not available in CPM. So, what is a
lag? Lag is a condition of waiting for a prescribed period before action can start. So, I
have given an example here. So, A is an activity, B is another activity; this takes 10 days
and this takes 8 days. So, this says; it is a start-start relationship. So as soon as A starts, B
can start. But I have given a condition called 5. So, as soon as A starts after 5 days only
B can start and I have to wait till the period is over, ok. And the negative condition is
primarily called as lead, which implies suppose I am writing it as minus 5; as soon as A
start, so B should have started at least 5 days in advance, ok. That is a meaning of
negative lag, which is otherwise called as lead, ok.
Now, multiple arrows and double headed arrows ok, primarily those are all not allowed.
The meaning of both is same. So, primarily I am not supposed to have two arrows in the
same arrowheads ok, which creates confusion when you are doing calculations, which is
generally not used. As I have explained in the last slide so, start-start, finish-finish are
generally used in combination in order to explain concurrency or partial concurrency of
activities, ok. There are some terminologies in the concurrency. So, one is called
complete concurrency; other one is called absolute complete concurrency, ok. So, this
complete concurrency is explained like this. I have an activity A, duration is 12; B,
duration is 7; as soon as A starts, B also starts; as soon as A finishes B also finish.
The logic is there is an activity with a larger duration and there is an activity with a
smaller duration; and the activity with the smaller duration is completely executed during
the course of an activity with the larger duration. It can have lead, lags whatever you
want you can keep it in complete concurrency. This is primarily an example of complete
concurrency. This case is absolute complete concurrency as the name absolute implies
the duration should be same; there is no lead, lag. So, exactly in the time frame of one
activity, the other activity is executed. It is primarily called complete, absolute complete
concurrency, ok.
Now next, in PDM loops are not allowed. So, what do you mean by loops? So, some
examples I have brought in up here. So, this is a start-start relationship, this is a finish-
finish relationship; but it is not like these two cases ok. In this case what happens, it just
goes in a cycle wise and these are all treated as a loops. In the computer programs, this
logic is generally not allowed. Now what about here, this is primarily here finish-start
relationship. This is a finish-to-start relationship again. So, two finish-start relationships
generally transform as a loop also. If you see here, this also is traversing like a cycle. So,
these relationships are generally not allowed in PDM. So, when you are working on
PDM, you have to see with combination of few activities that you are not landing up in
any loops, because loops are not allowed in PDM networks, ok.
So, what is the purpose of PDM and why do you want to have these relationships. Why
cannot I go ahead just with finish-start and why are we complicating the networks ok.
There is a purpose behind. So far in CPM networks we have understood, the there are
two types of relationships; one is a independent relationship, where in the two activity
does not have any relationship with each other, and the other category is primarily called
dependent relationship, wherein one activity is dependent on the other. For example, the
successor of an activity generally is dependent on the predecessor completion and so on;
that is primarily called dependent relationship.
Now, there is something else which is called semi-independent, ok. I need not have a
conventional finish-start, if I able to overlap the activities, ok. So, instead of waiting till
the end, why cannot I start my second activity, if the outputs for the first activity is really
generated from the same, ok. That is primarily represented a semi-independent and this is
not a forced overlapping; this is a natural overlap. There is an input required for B and it
can come somewhere not at the end, it can come somewhere in the mid execution of the
whole activity. Then I can start my B at that point of time, that is primarily called semi-
independent, two cases I have shown here.
Now, how do you represent this on PDM is, this can be represented with the start-start
plus a lag. So, if you want I can show this, so I having an activity A and I having an
activity B. So, I can show this with a lag, so maybe 5 or something, if this duration is
something as 5, then what I can do is after activity A starts after 5 days I can start my
activity B. I can easily represent this network on a PDM notation. The same thing I can
show it here. So, this is primarily by using a finish-finish relationship plus lead values.
So, I am having an activity A and this is my activity B and this I can use a finish-finish
relationship, this I can use as a finish-start relationship and I can put a lag here ok.
So, this can be finish-start plus lag ok. In the case of lag, I can explain how much time of
B has to be executed and then B can start or I can do is, I can have the same activity A, I
can have this activity B and I can put this finish-finish; but I can show you a lead ok. So,
minus some number maybe 4, here I am representing maybe 2. So, this can be 2 and this
can be 4. So, before 4 days of B completion, you have to activity A has to be completed,
that is a meaning of this. And this is after my start-start relationship, 2 days after then I
can start my B. So, either I can use finish-start plus lag or I can use finish-finish plus
lead. Both the cases are available here and both options are available.
So, it is primarily based on, what is the information or what is the relationship I am
having between A and B and what type of functionalities I can use. So, with the
combination of the relationships, easily you can show overlaps in PDM. That is a one of
the main purpose of PDM networks.
Now, I will just show an other example just to showcase the other purpose of PDM also.
This is an example as the; this is the just a trimmed example from a bigger example and
how do I read this example. There are primarily three towers ok, the real example had
three towers, tower in the sense it was an apartment construction and the three towers
had series of activities like brickwork, joinery, floor, painting and so on and they had
series of resources and etc.
So, I just used a short form for all these activities. So, for all the tower ones, I used 1 and
for all the tower two’s, I used 2 here ok, that you can see with the last number and the B
represents for Brickwork, J for Joinery, F for Flooring, P for Painting, same thing I have
used. And duration for all these activities I assumed based on the data I have obtained,
90, 75 and so on, and I repeated the same values for the 2nd tower, 3rd tower, because it
was exactly the same towers ok. And the relationship between all these activities are also
marked here ok
Now, how do I represent this? Suppose if you take a conventional CPM network so, how
do I represent. After brickwork in tower 1, I have to do joinery in tower 1 and same time
I can start my brickwork in tower 2, sorry so, this is a mistake. So, in tower 2 also we can
start, ok.
Now, same thing after my joinery work in tower 1 is done, so, after the brickwork in
tower 2 is also done, I can start my joinery in tower 2 and so on. That is how this activity
representation is generally done. Suppose if I want to represent this on a PDM network.
This creates little boring when you are having, let us assume 50 activities 10 times
repetition; then it looks really boring for somebody to really see. So, how do I make it
interesting in PDM? I can have a brickwork activity here. This is a joinery activity and, I
am just showing only two cases here. Total duration of brickwork is now 180 because of
90 and 90, total time taken for joinery is 75 and 75 fo r two towers; I am keeping it as
150.
So, how do I now fix the relationships; so, all these relationships I have, how do I fix it?
So, what I am going to do is, after the brickwork started 90 days after the brickwork,
joinery should start. So, this will cover my brickwork, after the brickwork this logic is
primarily covered. After the brickwork 90 days later, brick work takes generally 90 days;
then the joinery starts and this duration is given here with the lag ok. Now same case;
suppose if I do not put a finish-finish, I can do my brickwork and joinery may be also in
parallel which is not acceptable in this particular case.
So, what I have to do, I have to finish my brickwork at least before 75 days. So, the 75
days is primarily coming from my joinery. So, which implies which is also settling with
my finish-finish; so brickwork is fully completed and the last joinery maybe there is a 75
days of lag in order to finish my joinery. So, once I am fixing my start-start relationship
and finish-finish relationship, the complete representation on both the activities is really
done and this way you may have to extend for flooring, you may have to extend for
painting and so on ok. And whatever may be the other activities you may have to extend,
ok.
So, this is a main purpose of; the second main purpose of working on a PDM network.
First primarily to show overlaps and the next we can also show repetition activities in a
very nice way with the help of the other types of relationships which we have in PDM,
ok.
Now, then I have also been telling you always you should use the ends of the rectangle
alone for notations and there is a different meaning altogether, if you are using some
arrow marks from the either of the horizontal segments of the rectangle. As I have seen it
here.
So, this I have an activity A duration is 10, I have an activity B duration is 7 and this
arrow primarily is coming from the mid portion in a horizontal segment of activity A and
there is a 0.4 return. This implies after 40 percent of completion of A, I have to start my
B because it is connecting with the start relationship ok; that is the meaning of this. So,
after 40 percent in the sense it implies; after 4 days of A, I can start my B; that is what is
represented here. Suppose if A is duration is extending maybe to 20, then accordingly
this start of this activity B also you can exchange, you can vary ok. But what happens if
you do not use this percent complete approach, what happens and if I just use start-start
plus 4 then whatever if the duration of activity also changes; you may still go with the
deterministic value of 4 in any case, that you can avoid when you are using a percent
complete approach. The same thing for finish also I can do.
So, I am having an activity A of duration 2, I am having an activity B of duration 10. So,

once A is completed, then there is a finish notation to B and that is governed by 0.1
which implies; so, before that 1 percent of B completion I have to finish my A ok. So, 10
days is a duration of B so, 1 percent completion is actually 1 day. So, on the 9th day of
B, so A has to be completed on the 9th day of B; that is a meaning of representing this.
Suppose if the duration of A changes or B changes accordingly, the values also change
and this is an other usage on PDM network.
So, I have shown you what is a purpose of going with a PDM network compared to the
conventional CPM is fair enough and it works well in all situations. But there are
situations we wanted like overlaps, repetitions or percent completes, then we may have
to go with the PDM and it works well with the other types of relationships and with the
lead, lag combinations also, ok.
Now, before coming to that I said there is a prescribed time of waiting and so on; we
have very nice examples for all these. For example, in concreting, curing is a very
critical task ok; and instead of showing curing as one activity and then connecting it with
the finish-start relationship and so on. If you add the duration for curing then with the lag
ok; then that shows that you know it is taken care of and it is really understood by the
experts as to you have to leave 10 days for curing and so on, that is understood. So, that
is the meaning of all these relationships on the lead, lag combinations, ok.
So now, you understood how to represent a PDM network and what are the situations
under which you can develop a PDM network to really get the benefits of the PDM, so
all these we have seen so far. Now, how do I get the relationships and where are these
relationships coming in and duration for each activity I have shown in the last class on
CPM, how do we get the durations for each of these activities, ok. So, in this class we
will see, how do I get the relationships and how are these relationships coming in; little
you would have understood with the examples I have shown right now, but more detail
we will see.
This is a running example which I am taking from the first class; this is on developing a
computer laboratory, computational laboratory ok. Now after 1 or 2 iterations just to
develop and schedule. So, this was a real schedules which were developed after all
decisions were made, as to cupboards should be made by carpenters customized in the
lab and then furnitures should also be customized with the design, drawings and so on
and tasks to be made by the carpenters. So, like this several decisions were taken and the
durations also had come with that, some of the relationships also come with the help of
the decisions you made on each of the activities. For example, if the furniture has to be
ordered then the predecessors, successors would have been different, this furniture’s has
to be customized, sometime’s the relationships also change with these predecessors and
successors, ok.
Now, these are the list of activities for example; internal wiring that was a first activity
that was planned, after the room was partially ready. So, internal wiring generally it takes
2 days, so that all the wirings are hidden inside the rooms. And then next to furniture
cupboards because that was like customized option for us and to know what is a height
and to fix the full length of the cupboards ok. So, the cupboard was planned till the
ceiling height and it was planned to be done first.
So, after the wiring was done all along the room, then the cupboards were fixed, so that
the cupboards were like; the cupboards were planned to be fixed. Then flooring was done
and simultaneously false ceiling was also done, because they were not dependent on each
other and that could have been done in parallel. Then after finishing the flooring and
false ceiling in the remaining places, POP and painting was done on the same, so which
took 4 days. Then furniture procurement for tables and chairs were done and here this is
like a parallel task; and what is given here is this should be finished as soon as the POP
and painting are done, as soon as the room is ready the plaster of Paris and the painting is
completely done, then the table chairs can come into the room ok; that was because we
need a place to store all these inventory.
The next installation, primarily the customizing and assembling of the tables and chairs
were planned. So, that actually happens with the predecessor on procurement and also
with the electrical fixtures on, because tables cannot be moved, because we need power
points for all the tables, because we are planning to use a computer there. So, that is
where we had a link with the predecessors with electrical fixtures. Then hardware
procurement on PC’s and computers, for which we need the POP and painting because
once the room is completely ready then only we can bring in the equipment. And then
software procurement; so software can go in parallel, but the software procurement can
go in parallel in order to get permissions, approvals (Refer Time: 21:14) quotations and
so on.
But the installation if you see it depends on the room has to be done, initial machine
installations all has to be done, then only the software procurement also has to be done
for the installation to happen that is what you have I, K and L as a relationship ok. So,
that is how the relationship starts filling up, after you make all the decisions on the
activities. So, the logic is, let me go in little backwards the logic is first you have to go
with the levels, breakdown the project, as and when you start breaking down and moving
ahead you get more and more information on the assumptions, what has to be done; the
real scope of the entire work on each activity gets defined and then you go on in the
levels, ok.
So, at one stage all decisions on all of the items would have been done, then the network
schedule is generally prepared, ok. So, this is primarily done after the complete example,
after the complete what is it called a laboratory setting up is done. So, I may have a real
clean network schedules done, but it does not happen in practice when you are taking up
a live project ok. But this is only to showcase how to do the example, so I have used this
example. Now for the same example I have this, for me I need only the ID’s and the
IPA’s. So, I am, because I want to show you only how to make the PDM network ok.
So, I am taking this table which contains activity ID’s and the IPA, I have a list of
activities here and the list of IPA’s which I got from the earlier example are also here.
So, first activity A as seen here, there is no IPA and here in this PDM network you need
not follow the logic on left to right of arrow marks that you will see here, arrows will
come and go like anyway, ok. So, this is activity B, so there is nothing given on the
relationship type, if no information is given on the relationship type the default
relationship is primarily the finish-start. So, I am using finish-to-start on activity B, then
I have activity C, again no special information is given, so it is default finish start. Then I
have activity D, then activity E is primarily finish start with C and D, then I have my
activity F ok; activity F it is Finish-Finish with activity E then I have, for activity G, I
need to have 2 predecessors and J, I have not drawn so far.
So, I am now going down to the line, wherein I does not have an IPA’s so I am going to
start from I. So, I have finished till F, now I am starting with, so all these are done. So, I
am going to start with I again, I does not have an IPA; the next is J, so J has an IPA with
E, so this is a finish-start default relationship. Then I have H, H is also with E, finish-
finish with H ok, so this is also done. Then now I am going to do activity G because I got
the predecessors right now, so it is F and J are my predecessors for activity G; so I have
drawn this here. The next I have activity K, so it is primarily activity H and G; then I
have L. So, L is Finish-Finish with respect to K, so I have drawn here. Then I have
activity M, so it is primarily I, K and L ok; and then I have activity N.
So, after this I have activity O, I have used a cross line arrow marks here, with this is
with the predecessor is with E ok, this is also wrong. So, it has to come from till the ok, it
should not come in the middle, it has to come to the end of this box, ok. So, this is how I
draw my network diagram. So, if you see here, follow some pattern and avoid, as far as
possible try to avoid crossing over of arrows, this O could have been drawn here and
leaving a space on either sides, this O could have been drawn here, so that this crossing
over is not can be avoided ok, that is one suggestion.
And other thing is here you will see lot of dangling activities are there, this is a start
dangling, this is a start dangling, this is a start dangling which implies there is a lot of
time or float for all these activities. And if you see, this is a finish dangling activity, this
is a finished dangling, this is also a finished tangling activity ok; otherwise most of the
other activities are all connected. Now, you should be very careful with the dangling
activities because they may take lot of floats, ok.
So, this is how the PDM network is drawn, I will show you other examples also. Now let
us move on into network analysis, as I have told you while starting with the lecture.
There are two ways of network analysis; one is for assuming that the activities execute
continuously without any break once started, in the other which is called contiguous
activity execution.
In the other case the activity will have an interruption in the middle and it need not
continuously be executed till the completion ok. There are two ways, accordingly the
calculations on critical path also varies in both the cases ok. Now, this is a simple
example I have I am going to use, hypothetical example, so that I can show all the
relationships and also I can explain, so I took a hypothetical case.
I have seven activities here, activity A duration 20, no IPA and hence no relationship;
activity B 15 days of duration, activity IPA is A; C 10 A SS 2. So, this mean there is;
sometimes if I have a different activity here for example, F or something; then there is a
special relationship holds good only with the F activity ok; and in this case since nothing
is given, so this is special relationship is only for the activity A. When nothing is given
for example, like activity C then the default finish-start is always there ok; here also is
the Finish- Start that is understood. And this is finish-start plus lag of 5, this is lead 10,
this is lag 2, this is lag 4. So, these are the relationship I have used for my hypothetical
case.
The same example I have shown here so that, it is easy for you to follow how to draw the
network. So, I am having activity A which does not have any IPA at all A and duration is
20, so this is a network representation. Along with the critical with the network analysis,
I will show you how to make use of the boxes. Right now, I am only using my activity
name which is written on the top and the duration is written in the bottom corner ok, that
is how I am representing. So, this is my activity A ok, duration is 20; now next is activity
B, I am going to use this is a finish, this is a start relationship and the duration for B is 15
I am keeping it here. The next is activity C this is starts-Start and with the lag of 2 and
with activity A and the duration for C is 10 ok. Next activity D I have finish-finish
relationship with A with the lag of 2 that is shown here and there is a default C also. So,
which implies default finish-start relationship with C is also there ok.
Next is activity E; activity E has a predecessor with A and the relationship is finish-start,
but there is a special condition given on 5 days of lag ok. Then I have activity F; activity
F is primarily finish-start with E, but a lead of 10 ok; then I have activity G, default
finish-start with B, then I have finish-start with E with the lag of 2, finish-finish with F
with a lag of 4. So, this is how I have to represent my network. Sometimes, so I may also
have an activity name coming in here, then I have to use this combination only for that
activity; and if there is something else given here, then that default relationship also has
to be taken, ok.
Never try to overdo the relationships, also if no if see whether special type of
relationship is given, then that only has to take the dominant role. If no special type of
relationship is given, then the default finish-start will be the default relationship for the
particular IPA; that you have to keep in mind ok.

So, now the network is drawn ok, now we will work out for contiguous activity
execution, ok. Now contiguous activity execution it is same like your CPM network
analysis on AON. Primarily you have to calculate early start, early finish, late start, late
finish and early start, early finish for the forward pass, late start, late finish for the
backward pass and you will be calculating total float, ok.
And one thing you should keep in mind is and in contiguous activity execution and
activity once starts has to complete till the end and here you may have different types of
relationships. So, you have to take care of the different types of relationships and see that
the activity is not given a break in the middle ok. So, for example, I will just show this;
so, this is my early start and this is my early finish; that is a notation I have used.
So, I am starting with A because I know A was the predecessor ok. So, this is 0, 0 plus
20 is 20 ok so, 0 plus 2 is 2 here, 2 plus 10 is 12 here. I can also do this in parallel. So,
this is 20, so 20 is the starting, 20 plus 15 is 35 ok. Now this arrow goes till G, so I have
so many activities in the middle. So, I will just go with D now, now D. for activity D this
is 12, so it is 12; but what happens is, this takes a duration of this takes this stops at 12, A
activity stops at 20, 20 plus 2 is 22 so, this D has to finish by 22.
If D has to finish by 22, if I am going to go with 12 what will happen, I may have a break
in the middle because my duration of D is 7. So, what I have to do is retain the 22, 22
minus 7 becomes 15 here; that is a reason of keeping 15 here because activity should not
have a break in the middle. If I am putting in this has 12 then D can start have a break or
D can have a delayed start or D can finish early and then stop at 20. So, those are all not
acceptable in a contiguous execution ok. And now 22 plus 5 is 27 for early start of E, the
earliest I can finish E will be 27 plus 9 is 36, and now 36 plus 2 is 38 and this is 35
maximum is 38; so, I am going to go with 38 ok.
Now, 36 minus 10 is 26 use with the signs, so that is easy for you, 26 plus 8 is 34. Now
34 plus 4 is 38, 38 plus 4 is 42 so, the maximum is 42 and we are stopping it at 42. Just
go with the CPM network we used, early starts early finish same logic; only thing is
because of the type of relationships. You have to see to that we are not allowing a break
in the activities, under the same time we are matching up with the activity durations also;
that we have to keep in mind.
Now, backward pass so same logic here also. So, I am having here activity ok, this is
duration, this is early start, early finish from forward pass I have done, late start late
finish from a backward pass I am going to do ok. So, I am going to stop, I know that I
stopped at G, so I am going to start from here. So, this is now 42, 42 minus 4 is 38, so
this is 38; 42 minus 4 is 38, 38 minus 8 is 30 here, ok. So, this is 38 minus 2 this is 36
and this is 30 minus of minus 10 is 40 actually. So, the minimum only we have to take
into consideration. So, this become 36 and 36 minus 9 is 27 ok. Now this is 38, 38 minus
15 is 23 so, this is 23; 27 minus 5 is 22 here, this is 22 minus 2 is 20 that also we can
hold it up here ok; 22 minus 7 is 15. So, this is tally then 15 minus 10 is 5. So, this till
here we have done.
Now, coming back to this, so this is 22 minus 2 is 20, 23 this is 23 minimum is 20. So, I
am going to use 20; 20 minus 20 is 0 and this is 5 minus 2 is 3. So, the minimum value is
3 ok. You have to take care of the relationships also and you have to take care of the
durations also, which implies here itself I will explain one more time; 20 minus 20 is 0
here ok, as a result of the duration of the activity. So, this becomes 0, with the help of the
predecessor relationships this becomes 5 minus 2 which is 3, ok.
The minimum we have to take for the backward pass I am going to use 0 here. So, you
have to see for all the relationships and the activity also when we are working on the
backward pass. So, we have completed forward pass, we have also completed the
backward pass.
Now, total float for the all activities; total float is same like a earlier case which is
nothing, but you are your late start minus early start or late finish minus early finish so,
anything is fine. So, what will happen is some, so late start minus early start is 3 or late
finish minus early finish is 3; whatever you want you can try. So, for activity B it is 3, 20
minus 20 is 0, 5 minus 2 is 3, 15 minus 15 is 0, 27 minus 27 is 0, 38 minus 38 right is 0,
30 minus 26 is 4 here ok; this talks about total float of an activity.
If you see here what are all the activities which have 0 as a total float I have activity A, I
have activity D, I have activity E and I have activity G ok.
So, the critical path is something like this activity A, then I have activity E and I have
activity G. So, all these four activities are on the critical path. So, that is a meaning of
drawing this contiguous activity execution ok.
Now, let us move on to caution with the PDM, when you are working with the
contiguous activity execution then there are some problems with the finding on the
critical activities ok, just with the relationships. This problem will not happen when you
are working on interruptible activity executions, ok. Now few things I have brought it up
and like this there are so many cautions are there available ok; I am just going to
compare with normal critical, neutral critical and the reverse critical ok.
Normal critical you know what is it, suppose I am having a critical path and one of the
activity in the critical path if I am exceeding the duration, the entire critical path will be
reduced to one day of the duration ok, project duration will be reduced to one day.
Suppose I am taking one activity in the critical path and I am just decreasing it’s
duration, the entire critical path will have a one day decreasing duration. So, whatever
changes you make on the critical activities, the project duration will accordingly you
know show it’s reflection. If you are reducing it 1 day, project duration also will reduce
to 1 day; if you are increasing activity duration 1 day, project duration will also get
increased to 1 day, that is what is a normal critical, ok.
Now what, that is what I have written it here and how does a normal critical relationships
are in the PDM. Any relationship can end on. So, this activities are called the other
normal critical activities. The relationship will have and start it will have some
relationships and there may be some relationships which are continuing from the end
portion ok, these all will have normal critical. And what; when the activity duration is
decreased by 1 day project duration also decreases by 1 day; the activity durations
increase by 1 day the project duration increase by 1 day. This you would know in CPM,
it works well for all the normal situations ok.
There is something called neutral critical and reverse critical which is what I have
written here. Reverse critical this refers to an activity for which the project duration
shifts in the reverse direction to the shift in duration of activity. If I am decreasing the
activity duration by 1 day, project duration is increasing to 1 day; If I am increasing the
duration of activity to 1 day, the project duration gets decreased to 1 day, ok. So, when
you are scheduling the network and you are assigning the relationships as I shown in the
previous figures and all, you can have combinations of relationships; either I can show
finish start plus lag or I can use finish-finish plus lead that is, what is the example I have
shown.
So, when you are trying to map the relationships you should be very careful on or you
are landing up in all these cases, ok. That you have to be caution and this analysis on
contiguous execution only you have to be careful; in interruptible you would not have
any issues with the same ok. This is a neutral; neutral what happens is all the
relationships will end on start or finish and those relationships will also start from the
start or finish ok; it is like one sided relationship. Then what will happen in neutral, there
is no change at all when you are increasing these activity duration, if you are increasing
it to 1 day, there is no change at all to the project duration, ok.
In reverse critical the relationship has to be you know, the predecessor has to be on
finished side and the starting the successors relationships has to be on the start side of the
activity. Then these you will have a reverse critical type of an relationship, when the
activity duration is decreased to 1 day, the complete project duration will be increased to
1 day; when the activity duration is increased by 1 day here, then the complete project
duration will be decreased to 1 day, you can try with any example later on ok.
So, you have to be very careful with the contiguous activity execution, only on certain
characteristics you have to be very careful. Like this we have bi-critical, multiple critical;
there are so many types we have to be really cautious with the PDM analysis, ok.
The next is interruptible activity execution; this is little tedious type of a calculation. So,
some researchers said, when you are using a PDM network and having all the issues on
overlaps and so on; then there should be a sufficient number of breaks in between the
activities execution, otherwise there is no fun in developing a PDM network.
So, there is another group of researchers who really experimented on interruptible
execution and there is a series of formulas to calculate the critical path. There is a slight
difference in how do you calculate for interruptible. So, with these formulas it becomes
easy for you, I will also explained what is it. There is something called Task floats and
there is something called Relationship floats. So, you have to execute task floats also and
you have to execute relationship floats also, while you practice on more examples this
becomes very easy as well, ok.
So, I had a task start float. For example, I am having an activity ok, the start of the
activity alone is critical and I can start, I can give a break, I can continue or I can stop;
the finish or the completion is not critical that is primarily called the task start float ok. In
finish float, only the finish of the activity is critical and the start can have the real large
float; I can start on this day, I can start on this day, I can start on this day. Only thing is I
have whatever is the means, I have to finish the activity on so and so day because the
finish part of portion is really critical. In the tasks total float, I have to start the day of the
activity on a particular date, I have to also finish it on a particular date and the complete
activity is critical. Sometimes there can be a situation wherein, the start float is also
critical, the start is also critical, finish is also critical; but the entire activity need not be
critical ok, you will see situations later.
So, suppose if we want to have a total activity as critical; then that start-finish, total float
everything should be critical in the particular activity ok. So, now, let us see this task
start float it is nothing, but formula is late start minus early start. And what is the
meaning of the task float? The start of an activity or a task may be critical, even though
the task itself need not be critical ok, and task finish float which is given by the formula
as late finish minus early finish. The completion of a task may be critical, even then the
entire activity is not critical; it need not be critical ok. In task total float, the formula is
late finish minus early finish minus duration here then the activity that has a 0 total float
is primarily called a critical task. And an activity may also occur where-in the start is
critical, finish is critical; but the entire activity need not be critical ok, this is for the
activities.
Now since there are many relationships comprising of these activities, we should also
calculate where is the relationship floats available, ok. So, for each type of relationships,
I have there is a formula given here. For finish start relationship, the formula is late start
of the successor activity minus early finish of the activity minus lag; for start-start
relationship, late start of the successor minus early start of the predecessor minus lag; for
finish-finish, late finish of the successor minus the early finish of the predecessor minus
lag; for start finish, late finish of the successor minus early finish start of the predecessor
minus lag, that is how the formula is comprised in the relationship floats.
Now, let us start with the representations ok. How do I represent? So, divide the box into
9 zones ok. This is my activity, this is my duration as see earlier, this is early start of the
activity, early finish of the activity, this is late start of the activity, this is late finish on
the activity; start float is written here, the finish float is written here, total float of the
activity is written here. The relationship float is generally given by a triangle, because I
can also represent my lead or lag on the same arrows. To avoid no mismatches
relationship floats are generally written with the triangle symbol; and this is how the
representation goes on and the formulas are also written here the same formulas ok.
Now, the same example I have taken as explained earlier, the same example; so same
network I have right now.
Now, Forward Pass let us do for Interruptible Activity execution. So, what do you do in
forward pass? So, forward pass primarily calculating the early start and early finish of
the activities. So, I am going to start at A; so, this is 0, 0 plus 2 is 2, 2 plus 10 is 12 ok, 0
plus 20 is 20; and now this is there is no lead lag or something, no other arrows are also
coming, so, this becomes 20, 20 plus 15 is 35. Now let us start moving on here, this is 12
ok. I can still put that 12 because this is an interruptible activity execution; 12 plus 7 is
19 here ok, and this is 20 plus 2 is 22. Now what happens, the maximum only you have
to take; so now, this becomes 22; 22 plus 5 is 27, 27 plus 9 is 36. Now, if you see here 36
plus 2, 38 and this is 35. So, the maximum is 38. So, this becomes 38, 36 minus 10
becomes 26, 26 plus 8 is 34, again 34 plus 4 is 38, 38 plus 4 is 42; maximum is 42 and
this becomes 42 now, ok.
Let us do the Backward Pass. Now backward pass, I am starting from the primarily
calculating the late start and late finish of activities. So, this is 42, 42 minus 4 is 38 here,
38 minus 8 is 30 here, 42 minus 4 is 38, 38 minus 2 is 36 and 30 minus of minus 10 is 40
now. So, which is smaller? So, 36 I am going to keep this as 36, 36 minus 9 is 27 this is
here; this is 38, 38 minus 15 is 23, ok. Now 27 minus 5 is 22, 22 minus 2 is 20, 22 minus
7 is 15 this is 15, because I am having a finish start 15 minus 10 is 5 ok. Now 5 minus 2
is sorry 5 minus 2 is 3, 20 minus 20 is 0, minimum is 0. So, this becomes 0, ok.
Now, if I see here, now this relationship also I have; so, forgot 23. So, this becomes 23
this is 20, 22 minus 2 is 20, this is 23 only. So, minimum is 20 so, I have kept this as 20
ok; for this late finish of activity A ok. So, I have explained, so, backward pass is also
done. Now, just see this is wrong, this is a mistake ok.
Now we will calculate the total the task floats will start, ok. So, far we have seen only the
forward pass, backward pass; now task floats. The first task floats is start float of an
activity, primarily late start minus early start; where I am representing the late start, so
late start is here, the early start is here. 0 minus 0 is 0, 23 minus 20 is 3, 5 minus 2 is 3,
15 minus 3 is 12, see 15 minus 12 is 3, 27 minus 27 is 0, 38 minus 38 is 0 here, 30 minus
26 is 6. So, task floats are all calculated. So, if you see here start of A is critical, start of
E is critical, start of G is critical; because I have a floats of 0 values.

Now, let us do the finish float, a task finish floats, which is given by late finish minus
early finish. So, 20 minus 20 is 0, 38 minus 35 is 3, 36 minus 36 is 0, 15 minus 12 is 3,
22 minus 22 is 0, 38 minus 34 is 4, 42 minus 42 is 0. If you see their finish portions;
finish of A is critical, finish of E is critical, finish of D is critical and finish of G is
critical, ok.
Now, I am going to see the total task floats ok; task total float, late finish minus early
start minus duration. So, 20 minus 0 sorry, 20 minus 0 minus 20, so this becomes 0, 38
minus 20 minus 15, so this becomes 3 ok; 15 minus 2 minus 10, so this becomes 3, 22
minus 12 10 minus 7 this becomes 3. So, 36 minus 29 minus 9 becomes 0, 38 minus 24
minus 8 this becomes 4 and this becomes 0 here ok. So, if you see the total floats in the
activities, activity A is critical, activity E is critical activity, G is critical; only three
activities completely are critical.
So, how do I represent the floats, I have to still find out my relationship floats ok. Our
relationship floats, you have to be very careful and you have to look at the formulas for
each of the type of floats ok; and then you may have to work out.
So, this is my same network with all the task floats written here; start float, finish float,
total float in the tasks are all written here. And now first let us go with all relationships,
this is my finish to start relationship ok. So, this is finish to start. So, late start of this
activity is 23, ok. So, minus early finish of this activity is 20, minus the lag; there is no
lag value here, which is 0. So, this relationship float becomes 3 here ok. Now again I am
going to show with, I am going to show on start-start. So, this is start-start relationship.
So, late start of this activity, so primarily you have to take with the successor. So, I am
having a relationship. So, this becomes my successor and this becomes my predecessor;
this is represented in j form and this is represented in i form ok.
So, my arrow is stopping at here, this becomes my j now again this becomes my i for this
particular relationship ok, LS of j is late start of j is 5 minus early start of this activity 0
minus the lead or lag value is given as 2. So, this becomes 3 for me, which is written
here ok. Now let us see for finish-finish relationship ok; late finish is given as 22 ok,
minus early finish is given as 20, minus that is the lag value of 2 here ok; so this
relationship is primarily relationship is 0 here, ok. So, that is how it has to be done.
Now for let us do one more ok, let us take this relationship; this is primarily a finished
start relationship. Late start is 38 here ok, minus early finish of this activity is 35 minus
there is no lead or lag, so this becomes 3 here, ok. Now, let us do this, this is primarily
finished to again start only; but that is a lead value here ok. So, finish-to-start, so this is a
late start; late start is 30 ok, minus early finish of this activity is 36 ok, minus use the
same minus symbol and then you show ok. So, this is written as 30 minus 36 plus 10. So,
how much is this 40 minus 36 is 4, that is how you may have to work out ok. If you see
this relationship is having a 0 float, this relationship is having a 0 float, this relationship
is having a 0 float. If you have done the calculations perfectly fine you will obviously,
see a continuous link going on ok.
For example there is no break and you will see continuous stretch going on. If you see
here, this entire activity is critical because I am having a total task float as 0 ok. And this
relationship is critical, because I am having a relationship float as 0; only this finish
portion of this activity D is critical, because my this is 0. From here I am having this
relationship float also 0, so this relationship is also critical ok. And then I am having my
activity E. this entire activity E is critical; because my total task float is 0. Then this
relationship float is also 0, which we have worked it out; and now this complete activity
is critical because the total task float in G is also critical. If you see here you will follow
a continuous path throughout the network and you will easily find the critical path; if you
have done any calculation mistakes you will be easily able to find out in this interruptible
execution ok.
So, now so far we have seen two cases in PDM. So, one is contiguous activity execution
and the other one is interruptible activity execution. And this is one of the age old
method which people use for calculations; finding over the total the task floats, start
float, finish float, total float; then relationships, for each type of relationship use a
formula and then you will get relationship floats; and then you have to find out the
critical path of the whole project.
So, this takes 42 days for completion and you will see everything else is same ok.
So, now there is an other group of researchers also who have worked out on interruptible
activity execution, ok. So, the method of calculation is different, but the answers you will
get almost the same ok. So, they have used float formulas only for the activity, there is
no relationship or something worked out; and they have defined the floats in a different
way ok. So, one is called start restricted float defined as SRF; the next is finished
restricted float defined as FRF; the next is double restricted float as DRF; and the next is
unrestricted float URF; and then they also have something called single restricted float
ok.
So, what is Start-restricted float, you will see the name similar to what we have done
earlier, but this method looks really simple compared to this, so I have taught you in this.
There is no harm, if you can work out on this method also; there is really no harm in
doing it. So, the amount of time we can delay the rest of the work in the activity, after it
has started on it’s critical start date, without delaying the entire project. So, this looks
something like the start critical only ok.
And next is FRF Finish-restricted float, it is the amount of time we can delay the start of
work on the activity, or the rest of it has started then finish on time without delaying the
entire project. Here for example, if FRF is 0, it implies the activity is finished critical.
For example, SRF 0, it implies start of the activity is critical ok. DRF Double-restricted
float, amount of time we have to delay the rest of the work in the activity after it has
started, then finish on time without delaying the entire project. Here only the start is
critical, finish is critical and the activity as such is not critical, ok.
The next is URF, Unrestricted float, amount of time we can delay all or part of the
activity without delaying the entire project, ok. This is primarily the float complete float
on the activities you can really get with the help of URF.
The next is Single Restricted float, you should have a Start-restricted float and Finish-
restricted float both are together, you can call it as a single restricted float, ok. The way
calculations are done is little different from how you do; but otherwise you get the
similar results as I told you earlier ok, that is all on PDM network.
So, if you look at from the beginning, we have discussed on how to represent a PDM
network diagram ok. So, primarily on notations, we have also seen the purpose of
building up a PDM network and then from moving on from there we have also discussed
the two broad ways of critical path calculations; one is on contiguous activity execution,
the other one is on interruptible activity execution, ok. And PDM generally uses AoN
only, you cannot use AoA and so in the all the examples also you will see the same thing
that is happening. I also explained you how to define the logic, I also cautioned you as to
what are all the issues may come up, when you are putting the relationships.
So, when you are putting the relationships see to that, you are not landing up in
something called neutral critical or reverse critical or something. So, be careful in how
you are representing the relationships, and then you should be able to build up a proper
network ok. As far as software is concerned CPM or PDM, MS project (Refer Slide
Time: 59:07) anything any software can handle all the four relationships, along with the
lead or lag constraints ok. So, thank you, that is all on PDM, we will see with another
new topic in the next class, bye.
Lecture - 05
Line of Balance Method
Topic for today’s lecture is Line of Balance Method. Line of balance method is not a
new method, it is past 50-60 years. It is in industry and it is available. People have so
much used line of balance method and we will see how to find the critical part of the line
of balance method and before that we will have a short introduction on what is this
repetitive projects all about or linear projects all about.
So Linear Projects; linear projects are the projects involving repetitive activities. It can
be vertically or it can be horizontally repetitive in nature. Vertical repetitions primarily it
is not linear. Vertical repetitions, For example, typical flows in a building. Horizontal
repetitions can be highway projects pipeline projects and so on, ok.
Then several names are available to these methods and nobody knows where is the
history and what is the origin of all these methods but there are n number of methods
which are available and people are at most randomly used most of these methods. A few
of them I have listed and the list is not complete list. Construction planning technique,
vertical production method, time location matrix models, time space scheduling method,
location space matrix, time versus distance diagrams, linear balance charts, velocity
diagrams, linear scheduling method and so on. So, there are so many names available to
these type of repetitive projects.
So, what do you mean by the repetitive projects and what is this LoB, all about say LoB
and history. I told you it is a very old method existing ever since 50 years or something.
It is very widely used in the construction industry even now for the execution of
repetitive projects.
For example, for many projects like mass housing, high rise buildings, tunnels, highways
etc people use this very widely and what is this representation on LoB and how do you
represent a project in a LOB? As I told you this is a method for repetition, ok. So, if there
are several activities which are repeating now when then, then this is a very good method
to show it on that timescale. So, if you see here and this is a timescale.
So, I have written this in days and this is the units I have written here. So, this is a
timescale, ok. I have just shown two activities here. So, this is activity C, this is activity
C and this is activity F. So, how do I represent this? So, primarily this is a early start or
the start of the segments and this is the finish of all the segments which implies I am
starting my activity C on day 2 and I am starting and I am finishing my activity C on day
4 for unit number 1 which implies for the first time execution I am doing this maybe if it
is for a pipeline project, maybe for one stretch I am starting my activity C on day 2 and
finishing it on day 4. On the 2nd unit I am starting it on day 4, finishing it on day 6 as the
name implies here ok. So, these are the start and the finish of the segments in that
particular unit, ok. That is what I have highlighted here.
So, for example this is unit 2. That is how we write. This is a short general notation for
the repetitions to be shown. If it is multiple floors in a building, you can call it as floors 1
2 3 4. If it is like a pipeline project, you can call stretch of 100 kilometers or something.
So, like that you can vary this units but unit is a common name given to the repetitions
and this can be a time-line. This is how the LoB is a two dimensional drawing, ok. Now
if you see here, so at any point of time you are drawing a vertical line, horizontal line.
So, it shows us our start and finish of that particular activity along that segment ok. Now
what about these lines?
So, if you want to follow the crew continuity, so then I can also follow the crew
continuity. For example this is the crew. So, generally default is one crew size in LoB
diagrams. So, I have one group which is after finishing my 1st unit it is going to my
second unit, finishing my 2nd unit and then it is going to the 3rd unit and then I finish
and then I go to the last unit and this is my last unit completion.
So, if I want to know for example there is a crew composition called crew H or
something. So, what is the time duration taken for a particular crew? It starts on day 7
and it is finishing on day 19, ok. So, day 19; so, that is how I have to interpret the crew
representations also. Everything can be easily understood in LoB diagram, ok. Then,
next suppose if I draw a vertical line on any point of time.
So, for example I have drawn this on 11 12 13th day, I have drawn a line. So, this shows
what is the progress of the activities. For example, activity C has been completely done,
activity F has been completed till unit two and it has started the work has just been
started in unit number 3 and the remaining there are two units to be completed after this.
So, primarily it shows the status update or progress update.
When you are drawing a vertical line anywhere you draw a vertical line, you will come
to know what is the progress of the activity completions ok, then the next is a slope and
symbol I have drawn here. So, the meaning of the slope is. So this I have drawn ok. I do
not know whether you can see. So, this slope symbol that shows the rate of progress of
the activities.
So, if the activity is primarily close to the horizontal, it means a slower rate of progress if
the bar primarily the bar which I have drawn here now for activity seen in green color
activity or in blue color, if it is just close to the vertical, it means it has a fast progress or
fast rate of progression, ok. So, which implies I may have multiple crew. When I have
more than one crew, obviously the angle can be more ok. It starts tilting towards the
vertical, then it means the rate of progress for the entire activity has become progressed
very much, ok.
So, now let us take a case example to really understand what is the benefit of drawing a
LoB diagram. We have so many methods why should we keep on learning. So many
methods in scheduling that question can come to our mind.
So, for that the same example which I have showed you in the last class, I have just
brought it up again.
So, this is an example I have shown for discussing on the PDM example, ok. This you
would remember right now. So, there are so many activities like brickworks, joinery,
flooring, painting and so on. In two towers we have taken for example and only for two
activities or so I have shown you how to represent it on a PDM, ok. CPM obviously is
very boring because you may have to show all the activities repeating.
For example I may have 10 units, then all these activity will be coming in 10 unit in 10
times which is really boring and in case of CPM which is we call it as an activity based
and more an abstract based you will call the representation in PDM but what happens
they you will not even know when are the individual units completed started and so on.
Just with the 180 days you won't know when is the 1st unit completed, how many units
are there, how many units are remaining, units are more to be completed. So, it is very
difficult for you to understand in a PDM representation like this. So, next is a LoB
representation I have shown for the same example. Same units is shown here and days
are shown here. This example I have taken only two units.
So, I am only going to show the two units here. So, this is a real benefit of the LoB
diagram which is a line of balance diagram, ok. So, the first activity which is the
brickwork. So primarily you have to write it here. So, the brickwork starting on a day 0
and it is taking till 90 days here which is shown here and then the last unit 2nd unit is
starting on 19, it is completed till 180. So, that is how we represent as I have shown here.
So, this entire project is completed on approximately up at 400 something, ok. So, easily
you will interpret what is the different activities. It is not boring. You will also see what
is a progress update, how many crews are acquired, everything can be visually seen with
the help of a line of balance method, ok.
Now, with this example I have shown. Now what are the assumptions, how can I draw
this diagram. So, that is what we are going to learn in today’s class assumptions. Few
assumptions I have brought it up here, the logical CPM PDM network in each unit is
maintained which implies in this example if you see same example, I will just show you
here. There is something called IPA, ok. For each of these activities this IPA has to be
maintained for all the units in the in the network, ok.
For example, if you see here even in this LoB after this brickwork only, I am starting my
joinery works, then I am doing my flooring, then I am having my painting works. The
same thing is happening in unit number 2 walls. So, first brickwork, then joinery, then I
am having flooring activities and then I am going to do my painting activities.
So, the logic of the relationship in standard CPM notation or PDM notation what you
have followed it should be maintained in all the units. That is a number one condition
that you have to maintain. Next, crews work continuity is retained, which is also I have
told you. So, after the 1st unit the crew immediately has to move to the second unit.
Once it is completely done, immediately it has to move to the 3rd unit and so on.
If you have multiple crews also, the same thing has to be done ok. For example, I
brought multiple crew formations also here. So, this is primarily for single crew
continuity, this is for parallel crew continuity and this is for staggered crew continuity.
There are three patterns available, and generally what we do, we follow only a staggered
patterns in case of multiple crews and for single crews this is the only one pattern which
we have and parallel is keeping the two crews in one or maybe if you want to have three
crews, then keep all the three like this, then after then you start here and then you move
all the crews. So, primarily you are having something like a tower formations, ok. So, I
may have something like this and then this starts from here and then, I will have
something vertical and then it goes on.
So, this maintain this form of a parallel notations is difficult to know model or represent.
So, the assumption made in LoB is continuity we are maintaining, but we are
maintaining in a staggered form only. That is what we are doing in LoB.
The next is typical and non typical. What is this typical and non typical? With this figure
itself you will come to know. Durations. So, these are all the units and this is the
timescale, ok. So, now if you see here the duration of the particular activity in all the unit
is almost the same and that is also one of the assumption we have in LoB and but
realistically in real world you may have different durations in different units, ok. That is
not possible to be modeled in a normal LoB diagram, ok. So, generally we assume that
durations are the same.
So, for example if you want to do the same painting work, it may take a different
duration in one floor, it may take a different duration in other floor. There may be so
many reasons for same set of activities, but duration can vary at every activity for
different units at which you are working ok. But LoB does not assume all those. It
assumes only to be in uniform duration, ok.
The next thing in LoB is the buffer. So, what happens in buffer and why should we
introduce a buffer? Buffer is only for avoiding collision between the activities or we can
say interference between the activities. Suppose if you do not introduce a buffer, you can
see here the two activities are actually you know colliding with each other ok. There is
something called collisions, ok. Primarily you need time for preparatory works or for
setting up all your equipments and so on.
So, there should be a gap between these two activities, so that it is not really colliding
with each other. For example, this is primarily in the start buffer, this is primarily we call
it as an end buffer. Terminologies you will see later when we are working out on the
problems ok. So, we should have a little gap between these two activities, so that they are
not colliding or interfering with each other.
Now, let us move on for an example problem to really understand how to really do the
critical path or calculating or drawing an LoB diagram on a for a simple example, ok. So,
this is primarily an example on how do we do the interior for 4 classrooms, ok. We have
taken for four classrooms, all are four identical classrooms. So, how do we do the
interior works ok? So, these are the list of activities.
So, I have A,B,C,D, E, F till I activity and few activities only I have taken into
consideration. One is floor level marking, fixtures for roof panels and AC ducts, then
internal wiring, pipe laying for sprinklers, fixing roof panels, fixing wall panels, flooring,
painting and finishing and electrical fixtures. Duration for all these activities for one unit
we have worked out with one crew only which is taken as 4 hours 8 hours 16 hours 8, 20
24 and so on. The duration is as given here. Since hours is little difficult, we are
converting into day. The assumption for conversion is what? The default working hours
of 8 hours per day we have taken up and assumed to be assumed to be asked like this, ok.
Duration in day we have taken. IPA for all these activities are here, almost all activities
are in sequence except there are two parallel activities which implies this activity C and
D are supposed to be in parallel, ok. Otherwise all the activities are in sequence. And 0.5
days buffer is given between each activities to avoid the interference between the
activities. So, this is a sample data we have taken for showing the LoB calculations, ok.
Now, first step is to do a PDM network on one unit. Why we cannot do a CPM network?
It is because we have to show the buffers. The buffers are generally like a lag ok. If you
want to show it on CPM, you may have to unnecessarily add an activity just to showcase
a buffer. So, we are avoiding that. And it is better to go with the PDM network for one
unit, ok. So, same activity A B, then I have C, then I have D. So, these two are in
parallel, then I have E, then F G H I. As I told you all are in sequence except the C and D
are in parallel and everywhere I have used the 0.5 buffer between the activities, ok. Here
also 0.5 0.5; I have used.
And the calculations are done for forward pass and backward pass like in the last class
and the project duration for one unit is coming out as 19 days, ok. So, just for the one
unit if there is no break or something, so you will finish the entire one unit in 19 days.
Now let us work on the LoB calculations, ok.
So, this is little a complex a table. So, if you follow along with me, then I think it is very
easy for you to proceed. So, first what you do is forget about all the filled up data. So,
just you have to take from your previous table on the activity ID, IPA for each of the
activities, duration for all these activities, buffer, type, and duration. This generally I
have written it in the second line onwards but primarily it is assumed to be between the
two consecutive activities.
If you are if you are very comfortable in marking a line between these two activities,
then you can also start writing. So, primarily it has to be like this and you can write along
that ok, then time taken for n minus 1 units in this case n is equal to 4. So, this n minus 1
for us will become 3 units ok, then the schedule start date for the 1st unit and the nth unit
is what you have to mark it.
1st unit is 1st unit that same and the nth unit for us is primarily the fourth unit because
we are planning to work out for 4 classrooms, ok. So, the list of activities are taken from
the previous table A till I. IPA for all of them is also known, duration also I have copied
from the previous table only.
So, this is also there buffer type and duration cannot come right now. Only the duration
or the buffer value between two activities is given as 0.5 0.5 0.5. That I have marked
buffer type will not come. Now you have to start working it out, ok. So, this column is
completely done, this column is also completely done, this column is completely done
because you are filling up data from the previous table, this column also you are
completely taking from the data given in the problem only. The next is this. This you can
fill it up. Primarily what are you going to do is, this is primarily 0.5 into 3. So, time taken
for one unit is 0.5 and how many units I have? I have 3 units because n minus 1 is 4
minus 1 which is 3. So, this becomes 1.5.
Same way this is 1 into 3 is 3, 2 into 3 is 6, 1 into 3 is 3, 3 into 3 is 9, 3 3 are 9, 3 3 are 9,

2 3s are 6, 1 3 is 3. Like that you fill up this column. So, this column is also filled up ok,
then now we have to start filling in the remaining part of the table. So, first schedule start
day default we have to start on day 0 ok, then how do i. So, when do I start the 4th unit?
the 4th unit I will start after 3 units are completed that will be 0 plus 1 and a half which
becomes 0 plus 1 and a half. So, 1 and on 1.5 day you are going to start the 4th unit, ok.
So, that is what we have to write.
Now, I am going to see the second activity. Now for that I have to see the IPA also.
Second activity has an IPA with the first activity. So, I have to compare these two
activities which implies A and B from a rate of progress. How do I show the rate of
progress? The rate of progress since there are no multiple crews, the easy way to do is
you can see the durations and then you can see the rate of progress. This duration for
activity A is 0.5 duration, for activity B is 1.
Actually B is going little slower than A’s progress. Half a day, half a day, I am finishing
all my units in activity A. It takes one day for me to finish all my activities in activity B
so which will be completed faster. Obviously A will be done faster than B. So, I can
introduce this start buffer, ok.
So, I am going to introduce a start buffer. The minute you are going to introduce a start
buffer, then you have to start filling up the first unit, ok. So, what will be the scheduled
start date? This is 0 which is coming from here plus duration of activity A is coming
from here which is 0.5 plus buffer between the A and B is again another 0.5. That is a
next 0.5 which is coming here and this 0.5 is coming here. So, this becomes 1 for me and
when can I start the last unit. As usual, this is 1 and plus, I need 3 days to fill up, to finish
up all my 3 units. So, 1 plus 3 this becomes 4.
Now next activity is C. C has a sequential relationship or dependent relationship with B.

So, I have to look only for these two activities. What about the duration? This is 1; this is
2. So, it is still slower than the first activity. So, still I am going to say start buffer only.
The minute I say start buffer, I have to fill up my 1st unit only first and then I have to fill
up my 4th unit. So, this is again 1. It comes from here, this is 1 plus duration of activity
B is 1, plus 0.5 is the buffer between B and C that is also 0.5. This becomes 2.5. So,
schedule start date for the 4th unit will be 2.5 which is here and plus time taken for
finishing my n minus 1 unit is 6. So, this becomes now 8.5.
Now the critical issue starts which is parallel activities ok. Now I am having activity D
IPA is B. So, I should not look with C. I have to look with activity B only. So, what
about activity D and B? So, activity B and D both are having duration as 1. So, what
happens here? Start buffer, end buffer both are same because there is a same rate of
progress between these two activities. It is comfortable for me to write start buffers. So, I
am going to write start. If you want you can write end and then you can work backwards
also. There is no harm in it, ok. So, now what I am going to do here?
So, now this is I am I am stopping B at 1. So, I have to take 1. Do not by mistake copy
the 2.5 here. So, this is now 1, 1 plus duration of activity B is again 1, buffer is also 0.5
between these two activities. So, this is again 2.5 and this is this 2.5 plus 3. This becomes
5.5 here, ok. Now next is activity C.
So, for activity C, I have to check activity C and D. Both are my predecessors, ok. So,
now activity C has a duration of 3, C has 2 and this has 1. So obviously this has a still a
slower rate of progress. So, I am going to still introduce a start buffer only, ok. Now
what am I going to do? So, both these values are 2.5. So, I am going to start at 2.5 plus
duration of activity C is 2 which is coming from here. So, this is the 2 plus 0.5 is the
buffer between the D and D, E or C and E. Both are same. So, I am going to add this 0.5.
So, this becomes 5 for me, ok. Now here what happens this is 5 and the duration taken
for n minus 1 unit is 9. So, 5 plus 9 becomes 14 here, ok.
So, this is done now. Activity F it has an IPA with E. So, I have to compare activity E
and F. Only both are having same duration. So, again start or end buffers, both will have
the same results only here. So, I am going to introduce the default start buffer here. So,
this is now 5 plus duration of A is 3 here. So, this 3 is coming here buffer between E and
F is 0.5, this is 0.5 and this becomes 8.5, ok.
Now scheduled start date of 4th unit is so this is 8.5 plus 9 units it takes for n minus 1
units it takes, 9-9 days. So, this becomes plus 9. So, this becomes 17.5, ok. So, now
coming to the next activity G and IPA is F which is an immediate. So, I have to compare
these two activities. Again, these two activities are having the same duration.
So, I am again introducing a start buffer here, ok. Now this is 8.5 plus time taken for
activity F is 3, that 3 is coming here, plus 0.5 is the buffer between G and F that is 0.5.
So, this becomes 12 now ok. Now for the 4th unit this is 12 plus time taken for n minus 1
is 9. So, this becomes now 21, ok. Now I am at activity H, IPA is G. What about the
duration? I have to compare with H and G. This duration is 3, this duration is 2. So, what
happens here? So, this has a little smaller a duration than the previous one. So, we may
have to introduce an end buffer. So, we may have to introduce an end buffer here ok.
So, I have to start from the 4th unit only, ok. Listen carefully here. So, since the duration
is little smaller, I am introducing a end buffer. The minute I am introducing end buffer, I
may have to start from the last unit. So, this becomes 21 plus duration of G is 3, ok. That
is coming from here and plus buffer is between G and H is 0.5. So, this becomes 24.5.
How can I find out the scheduled rate of the start date of 1st unit? It is 24.5 minus 6. So,
this becomes 18.5, ok.
If by mistake if I am starting on 12 plus duration of activities, 3 plus 0.5 which is 15.5 as

by using a start buffer what will happen is after 2 or 3rd unit I may be just colliding or
interfering with the F activity or something.
So, in order to avoid that only we are introducing an end buffer and then we are deciding
the scheduled start date of H in order to avoid the interference between the activities. In
the next activity is I and the IPA is H and if you compare the H and I, this is 2 and I is
only 1. So, obviously it is a faster rate of progress and it will be finishing fast. So, again I
am going to introduce an end buffer. When I am introducing an end buffer I have to
come from the last.
This is 24.5 plus duration of H is 2, buffer between H and I is 0.5. So, this becomes 27.
So, what about the scheduled start date of 1st unit of ith activity? So, this will be 27
minus 3 which is 24, ok. Now what is the project duration? I am having project total
duration that is 27.
So, scheduled start date of 4th unit is 27 and plus 1 day duration of activity H you have
to add. So, 28 days it takes to complete for the entire 4 units, ok. Now that is what we
have to understand with the help of this LoB table. So, this table will help you to make
the LoB diagram.
Now, what I am going to do, I am going to copy my 1st unit values, 2nd unit value, so
that it is easy for me to show, it is easy for me to finish up the LoB diagram. So, I have
copied my table here, ok. All these values still 27. I have taken start data 1st unit and the
4th unit of all the activities I have obtained with the previous table, ok. So, now I am
going to plot the LoB diagram. Only for critical activities I am plotting right now.
So, the D is the non-critical activity which we have identified earlier, and which is also
very clear with the duration also. So, D is the non-critical. So, I am not showing it right
now. There is a purpose behind. I will show it later. So, same way so y axis I am having
the number of units. So, 1 2 3 4 units four classrooms x axis I am keeping days and the
duration is 27. So, I am having a unit till 30, ok.
Now what I need is these points will tell me what are the points to be plot on the graph.
Maybe you can take a graph sheet or you can plot it on excel using charts. Whatever
method you are comfortable, you can do. I have used charts for plotting the same. So, A
activity. 1st unit, start of first is 0, start of 4th unit is 1.5. So, this value is actually 1.5 and
this is 0 here ok.
Now what about the finish of activity A? 1st unit that will be 0 plus 0.5 adding up the
duration. So, that is 0.5 here. 1.5 plus 0.5 this is now 2, ok. Now first you got all the four
points. So, you can just plot the activity A. So, this is now activity A. Now I am starting
with activity B. The 1st unit is starting on 1, ok. I am having a 0.5 buffer, start buffer that
is why starting on 1 and the 4th unit is starting on 4, sorry is starting on 4 ok. So, this is
also here. So, this is 1 and this is 4, ok.
Adding duration of B. So, this is 1 plus 1 is 2, 4 plus 1 is 5, ok. So, now I got all the four
values. So, I can plot my activity B ok.
Now next activity C. So, activity C first unit. So, this value is 2.5. So, this value is 2.5,
and this value the start of 4th segments. So, this is now 8.5 adding the duration of
activity C. So, this becomes 10.5. So, adding plus 2. So, this is 2.5 plus 2, this is now 4.5.
So, now I have got my plots for my activity C ok. So, similarly we have to finish up for
the all the other activities. So, activity D, I have skipped right now because it is not
critical. There are other ways of showing D. We will show it later. Now activity E. So,
starting as 5 and 4, the last unit is on 14. So, 14 plus 3 this is now 17, ok. So, 5 plus 3
this is now 8, ok. So, activity E is all the four points are obtained. Activity F, So, this is
8.5 duration for F is 3. So, this is now 10.5. sorry this is now 11.5 and the nth unit is now
17.5. So, this is activity E and I am on activity F. So, this is now 17.5 duration of F is 3.
So, this becomes 20.5.
So, and I have got my bars. So, these three bars are in parallel because duration of the
three activities are same that you can see here 3 3 3. So, all these three activities you can
see E F and G are like parallel activities, ok. The bars are almost like parallel and since H
and I are having a reduced duration than E F and G you will see here. I have introduced
only n (Refer Time: 33:15) buffer in both the places. There is a lot of float in the starting
place in the start units 1 and 2 in both these activities, ok.
So, that is what you can infer from these diagrams. Apart from a horizontal line what is
the interfere meaning of the horizontal lines, any vertical line, What is the meaning? You
can still capture with the help of or the LoB diagrams, ok.
Now I am going to show all activities. There are two ways of showing all the activities.
One is I have kept this activity D, this is activity D. You can overlap the non-critical
activities also and merge along with the critical activities like how I have shown here, I
have overlapped this. So, this is my activity D. I allowed it to overlap with activity C
which is also possible that you can do or you can have two axis; one for critical. So,
these are all for critical and you can keep the non-critical activities below the axis and
you can still show the same all non critical activities down. So, both the ways are
possible. You can use any one of the options you are comfortable ok.
So, this is a very simple very simple case on how to represent a very typical duration on
representing it on LOB. All units are same, everything is same, duration for all activities
are same only. One crew we have used ok. Suppose if there is a pressure on completion
and I have to finish this activity for example same example, we will take ok. This entire
project is completing on 28 days.
Suppose if I have to finish the all the projects on day 15 or something if there is a
pressure on completion happens, then where all I should employ multiple crews and how
many crews I have to employ still how can I maintain the crew continuity, there are
calculations done which I am not covering in this particular lecture, ok.
But people have worked on how to use multiple crews in order to shorten the project
duration. So, this general assumption is I have used only one crew. You can see that you
can calculate. See here this is the first crew. The same crew will be moving up for 2nd
unit, the same crew is moving up for the 3rd unit and then it is moving up till the end for
the last unit. You can see that in all the activities, the default assumption is only one crew
is assigned for all these activities. That is where you have done, you have got the
calculations like that ok.
Now, so that I have not covered. Another segment of work also I have not covered, but I
will just introduce and complete this lecture which is called non-unit based. There is
something in the literature which talks about unit-based repetitive activities and non-unit-
based repetitive activities. So, what is the meaning of unit-based and non-unit based;
most repetitive scheduling methods available so far and assumed of identical sub units
and identified units of repetition. For example, even in the same, let us assume I have an
activity called Excavation sub base and so on. So, excavation may look same for the
entire stretch maybe drawing you are laying a pipeline of 240 kilometers or something
for the complete stretch in highway road or something.
So, for the complete stretch you may be having an activity called excavation, but the
entire sub unit may not be the same. I may be having a different type of soil. I may be
also landing upon rock excavation. So, the equipments may differ the way, number of
sub activities in order to do the excavation may differ. I may not be having the same
activity as such throughout in my sub units also throughout in my repetition stretch, ok.
So, they are all called non-unit based of repetitive projects, ok.
For example this is again I have taken two examples just to showcase, just to explain you
or make you understand what is a non-unit based. I think you have heard of highway
widening projects, 2 lanes to 4 lanes or 4 lanes to 6 lanes.
So, this is also a highway widening project if you see here. So, there is a proposed
central line here. So, the existing carriageways this color shows the existing carriageway.
This is the new carriage way, the violet ones are the new carriageway which is proposed
and the red one is primarily the proposed to central line. Earlier this is the central line for
here and this is a central line for here.
Now this is a proposed central line as a result of the widening of the road projects, ok.
This is primarily called right side widening. The same thing happens in left side
widening, ok. So, this is an existing one, this is a proposed one and this is a proposed
central line in case of left side widening.
Now, suppose if you want to have concentric widening then I may have the proposed
central line and I may have to widen the project on both the sides ok. That may happen in
the concentric widening, then the next is with service road left side widening with
service roads. So, there is an existing carriage-way here and then I may have a service
road connections in order to widen with the service roads.
Now what happens here in this particular cross section, suppose I am taking a highway
project, I have to do let us I say 250 kilometers or 300 kilometers of a stretch of National
highways. I am going to widen the highway in to 2 lanes to 4 lane or 4 lane to 6 lane
whatever is a case what happens throughout the entire stretch if you want to draw a LoB
diagram, you may be identifying few activities like For example weight mix macadam,
then granular base, sub granular base like that excavation embankment, you will have
similar set of activities throughout the entire stretch, but what happens the items or the
sub steps under each of these activities will not be the same because of the differences
you may have to go through in the project. Somewhere some place you may have to do
right side widening some place.
You may have to do left side widening, some place you may have to do concentric
widening and accordingly what happens the sub activities or the sub steps in each of
these activities will start differing and this is widely very famously called as non-unit
based repetitive projects.
Now, for the same example which I have shown now as the interior of the classrooms,
ok. I may be doing the interior of 4 classrooms maybe even 15 classrooms, I can even do
for 20 classrooms. If you want to showcase an LoB diagram, the activity items which I
have marked right now will be the same, but the number of sub steps will be different
and hence, duration can also be different for all of them.
For example, I may have different capacity lecture rooms. I have taken few pictures just
to show that the different room layouts itself can be very different. It is not just addition
of table chairs and so on but it can be even the layout inside can also be totally different,
ok. Here you do not have a projector or something. Only main projector with the fire
protection systems and so on.
This is the arrangement little bigger classrooms, two projectors, two whiteboards, one
blackboard. So, all the setups you will have more settings in the same room. So, a
different way of setting up all your fire sprinklers, your(Refer Time: 40:53) ducks
running up your AC services and so on.
So, this is very different, little more a larger a classroom. So, what happens is the items
may be different ok. I may have the same activities like what I showed you earlier like
floor marking, floor finishing and all these may be same but the number of units, sub
units in all may be varying. So, with this I will just stop my lecture on Line of Balance
Method.
So, today I have just introduced a small topic only which is how to calculate the critical
path and how to just showcase a LoB diagrams. That is what we have seen and that too
only for a regular case which is too simplistic in nature, only single crew on all activities
and with sequential or parallel activities in place; how shall I you know plot my diagram
that is all we have discussed but there is so many other things to cover which I am not
covering as part of this lecture. If you are interested you can go through relevant
literature and you can understand.
Thank you bye.

Lecture - 06
Resource-driven Scheduling
The title for today’s lecture is Resource-driven Scheduling ok. So, let us see what are the
contents I am going to cover in this class. Primarily, I will give a brief introduction on
what is resource scheduling.
So, resource allocation or resource loading and so on, and little extend to resource
leveling also you will understand in today’s class. Then, we will move on to RCPSP. So,
RCPSP is resource constraint project scheduling problem, introduction to that we will
see. Then, RDM representation, I am just going to cover there are two means for
resource scheduling: one is RDM representation which is relationship diagramming
method and we are going to see the other form of modes other form of RCPSP which is a
mode representations ok.
So, for all these cases I have shown you trial and error method on how to evaluate and
find out the answers, but in real practice if you are having a large number of activities in
a project it is very difficult to work out manually and work it out on trial and error
method. So, we need to have heuristic methods or other combinations in order to solve
these sort of problems ok, but we will understand broadly on what is this resource
schedules today.
So, what do you mean by resource? So, resource can include any physical variable
starting from man, material, machinery, sometimes even finance most of the time even
space can be a real big issue when you are planning on resource schedules ok. But, in
general what comes to peoples’ mind is whenever you are thinking of resource schedules
and resource leveling and so on, it is man and machinery only. Only manpower
machinery that is all, others do not play a major role because materials you are not
actually carried over in the long run. So, it is; so, it is not generally off very serious a
concern.
The next is resource allocation or resource loading; both the names are really same. It is
nothing but dumped in the resources on each activity to the required amount and time or
duration of each activity and no need to do anything on reducing the resource or over
allocating the resource and so on ok. Accordingly we have two types one is called frond-
end loading and the other one is called back-end loading ok.
Now, let us see a simple example. I have taken a hypothetical example; four activities are
there. Activity A and C are in parallel; I think now you know better on how to
understand and interpret this. B and D are the sequence steps number 2 in the list ok. So,
they are actually successors to A and C and the duration for all these are given, resource
required is given. The assumption is all these four activities share the same resource,
then only you can do all the allocation and loading together ok.
Now, let me draw this network. So, how do I do? There should be an arrow mark from
here ok. So, primarily this is my network. So, once A is done B is done, then after A and
C are done D is done as per the predecessors relationships, this is a network and, now, if
you see this network the resource allocation for all these activities are also given here
and I am just going to show the front-end loading. What happens in front-end loading?
Front-end loading is based on your forward pass ok.
If you do forward pass calculations for this let us assume I am only going to show on
these activities. This is 0, this is 6, this is 0 and this is 4 ok, this happens to be 8 and this
is now 6 and 8, this is now 4, 6 maximum is 6, 6 plus 5 is 11 ok. So, 11 days is a total
time taken for my project for this small project, and if you see these resource
combinations so, I am actually consuming a resource combination of this is my resource
consumption until till 5 I am occupying.
Now, if you see here the front-end loading says I am actually dumping in resources
based on my forward passes which implies yearly starts and the yearly finishes. So, as
such A starting on 0 finishing on 6 ok; now B is starting as soon as A is done, so, 6 till 8.
Now, activity C is starting on 0 until till 4 and D for D I have to wait for A and C to be
over. So, D is starting on 6 and finishing until 11 that is primarily as we call it as a front-
end loading resource aggregation chart ok. If you are working on a backward pass; so,
this is 11 now, this is also 11, 11 minus 5 is 6, 11 minus 2 is 9. Now, this is also 6 for me
6 minus 4 is 2. Now, this happens to be 6, 6 minus 6 is 0 ok.
Now, I am actually employing my resources based on my late starts and late finishes. So,
D is from 6 to 11, A is from 0 to 6, C is actually from 2 to 6 and B is also from 6 to 8 if
you dump in the resource based on the backward pass then primarily you call it as a
back-end loading of resource aggregation chart ok. So, this is the x-axis you have time in
days and y-axis you have number of resources in the x-axis. Both these cases duration is
finishing on 11 and the maximum resource taken up is 5 for both the cases ok.
So, primarily we call it as a resource allocation or loading. Again, there are so many
ways of showing it. You can also show tabular form, you can also do like this or you can
show day wise separately and even the order of these activities also need not be AD in
the bottom you can keep it on the top you can do whatever forms unless you are
representing the number on each days are same ok.
Now, let us discuss on resource scheduling. So, primarily there are two forms of resource
schedules one is called time limited, and other one is called resource limited. So, what
happens here is in resource scheduling why am I bringing in time every time because the
number of resources determines the duration or the minute you fix the duration you have
to decide on the resources. So, the resource and the time are actually interlinked with
each other ok.
So, now when you are doing on resource schedules primarily I am working on
scheduling part then there are two parameters which comes together; one is time
constraints or do I have resource constraints ok. So, the project in time limited issues the
project must be finished by a certain time using as few resources as possible. So, it does
not mean since I have to crash my, it is not crashing. So, primarily you are finishing in
the last example for instance I have to finish it on eleventh day ok.
So, but primarily you are going to finish the same project on eleventh day only with the
few resources as possible. I am not going to delay my whole time and then use more
resources and do something and so on. So, that is not primarily the whole concern, but it
is a time and not the resource you would say which is critical in this particular phase of
time limited constraints.
The resource limited case the project must be finished as soon as possible again there
should be no delays or extended duration on any of the activities etc but, without
exceeding the available resource constraints. So, in the earlier case maybe there was a
constraint on resource 5. So, I could finish the project on 5 and there was nothing
happened here ok.
Now, when there is no resource constraints and I have abundant resource with me then
both time-limited or resource-limited is not a problem for me ok. Even if I slip one or
two days still I could finish in the respective time or if I have more resources then
obviously, I am not going to worry about how know meticulously to finish my whole
project.
So, we both the issues will not be complicated at all when I have scarce resources, but in
reality what happens in construction site, the resources are always scarce in nature. It is
not abundant so, always your resource scheduling has a critical challenge there ok. So,
here also so, there is a mistake here ok.
So, now, this is the same example. I have brought in the same network diagram also. So,
now, what do you have to see is; so, this is time limited. So, you earlier we have seen we
have to finish the project on eleventh day. So, finishing the project on eleventh day or
tenth day, what happens is I am not disturbing anything here maybe I am using more
resources whatever it is, but still the critical time at which I have to finish I am meeting
that in place ok.
In the next case on resource limited again, I can finish the project on 5 with 5 resource
maximum or with 4 resource. So, primarily I am finishing that, but I am not extending
my duration from eleventh day to any other days. So, that is what is a whole meaning on
time limited and resource limited ok.
The next is resource leveling because whenever even in the last example you have seen
here there are so much of idle time for workers. Suppose, if you want to schedule this
project and then mobilize for the manpower, then you may have to call all the workers
for all these days and what happens and you may have to pay for idle days of the
workers. Suppose, if it is equipment you may have to pay the idle sitting time of all your
equipments in the site ok. We generally it is like too much of cost overrun and which is
actually a waste so, which you should avoid ok.
So, what is resource leveling? So, minimize the fluctuations do not call do not have an
platform in which one activity I am calling. I need 4 workers on one day, 10 workers on
7 on second day then 3 workers on the third day, then 11 workers on the fourth day
which is like too much of fluctuations I have in my whole project ok, seek to that am
having utmost like uniform 7, 6, 5, 7, 6, 5 something like that which is close to each
other which is an acceptable issue in the sites ok. So, resource leveling is minimizing the
fluctuations in day to day resource use throughout the whole project.
Now, why should I level the resource to improve the work efficiency and to minimize
the cost on the whole project. As I told you unnecessarily you may have to pay for the
resource by keeping the resources in idle ok; in order to also what happens is the
lethargic flow on one day and active work mode on the second day, again lethargic mode
active mode also will obviously, you know start spoiling your work efficiency or
productivity and so on. So, for all these issues better to have a uniform peak load of
workers on all the days of the project.
Can all resource can be leveled or all do resources has to be leveled. Not all resources,
there is no point in actually looking at materials and other resources on leveling. So, only
the labor and equipment or manpower and machinery are the only one resources which
you have to think of in leveling.
So, now, what are the assumptions you have to make? So, one assumption is difference
in productivity between the resources of same type or different type groups are not
considered. So, I may have a crew called and a carpentry crew, I may have a crew called
a masonry crew. The helpers in both the crews if you want to swap then you cannot think
of difference in productivity between resources between a manpower of different work
crews or within a masonry crew itself different workers can have different productivity
levels and accordingly you may have a different duration and so on. That also you cannot
consider when you are working on resource leveling problems.
The next is a real serious assumption. Activities cannot be split once the activity has
started. So, if you are planning to start the activity on day 0, so, finish the completely
loading the resources and you cannot stop the resource loading after 2 days and then
continue after 3 days and so on which is not possible in resource leveling. Now, let us
look at the same example ok. So, I have modified this example to show the fluctuations
how it is done in trial and error practices in sites.
Four activities, same example ok. So, this is my first combination 5 resources. So, lot of
know idle resources are here if you want to maintain the five fluctuations are there, some
days I need 5, some days I need 2 and some days I need only half of the crew and so on.
I need only one worker in the site and so on. So, some days I need 2 this is stands for 1,
this is primarily 2 and this is still 5. So, that you still wanted to avoid ok. So, what I am
going to do I am trying to minimize or level it at resource number 4 maximum
availability I am going to call is 4.
So, what happened since because you are trying to cut down this then the duration for C
and B will get extended because A and D is on the critical path. The general assumption
in resource leveling is do not touch the activities in the critical path and you will have to
always look at the non-critical activities in order to know minimize the resource and then
so that even if the duration is extended it is not delaying the entire project ok. So, what
happens the C, B are the non-critical activities I am looking at the non-critical activities
and I am actually extending the duration, so that my I am meeting my resource of 4.
What happens if I go below 4 and so on, then the duration will obviously, slightly get
extended which also we will see what sort of a problem it is. Now, for the same example
let us see a trial and error method of how to do the leveling problem because I showed
you with for small example I showed you, but there is a trial and error method. There are
also scientific methods called minimum moment algorithm which is available in order to
do the leveling exercise ok.
Now, for doing the leveling I need to know the early starts, early finish, late start and late
finish. So, for forward pass, backward pass I have to work out. So, I will just quickly
write this. So, this is 0 6, this is 0 4 here and this is 6 and 8 and this is my 6 and 11
working backwards. So, 11 minus 2 is 9. So, this is 11 minus 5 is 6. So, this is 6, 6 minus
4 is 2; this is 6 and 9 minimum is 6, 6 minus 6 is 0 ok.
Now, the critical path you obviously, know it is 6 and 5. So, this has to come here ok. So,
0 till 6 this activity A and activity D is 7 till A, 7 till 11 ok. So, now, this I have dumped
in because I have to know where is my path until which I can dump in the resources ok.
Now, here critical path obviously, you cannot do anything so, I am going to put the
resources. Resources for activity A is 2 on all these days. So, this becomes 2 and for
activity D it is it is 1. So, this is now 1, 1 on all these days ok.
Now, what about my activity B? My activity B, I have actually a float from 6 until 11
and for activity C I have a float from 0 until till what date? Until 6 because it starts from
early starts and late finish and also you should keep in mind you should not violate the
logical relationship of the network as well ok. Now, what you should do? So, primarily I
wanted to know what is my maximum crew composition, suppose if I want to increase or
decrease a resources what is a duration with which I have to extend; that obviously, you
have to understand ok.
So, for which you have to know what is the work content of all these activities. So, work
content or total quantum of work. So, this is 12, this is 8, this is again 12 and this is 5 by
multiplying the duration and resources ok. So, activity A is done and activity B is also
done now you have to work with only these two activities ok. Now, what is the option
available activity C? I have to finish how many 12 and so, what I am going to do instead
of using three resources suppose I have done 3 resources it will be 3, 3, 3 and so on
wherein you cannot do. So, what I am going to do I am going to use two resources on all
these days. So, 6 into 2 becomes 12 so, which is tallying my combination ok.
Now, we will look at activity B. So, activity B, I am having 8 work content day. So, I
have to finish the 8 on all these days. So, I am going to put 2 for few days and then I am
going to use the 1, 1 for the remaining days. So, 2, 4, 6, 7 and 8 so, this becomes now 8.
So, all my resources on all the days has been done. If you see here, so, this is 4, this is 4,
4, 3, 3, 3 and 2 and 2. This is not too much of fluctuation ok. It is actually uniformly
decreasing ok. So, this is acceptable, but do not have too much of variations like 4, 3, 4,
3 and so on which is not that good.
Now, if you see here what about my logic after A after A, B is starting the logic is
maintained and after C after C and A my activity D is starting. So, both the logics are
actually maintained and there is no problem in dumping this resources. So, you have to
look at the logic, you have to look at the work content has been settled. So, this is 12, this
is 8, this is again 12, this is 5. So, that work content is also now settled. So, this is
primarily a leveled diagram.
If you want to look at other combinations I think it is not possible either. What other
option I can say is, you can keep this as 12 this becomes 0. So, this still becomes 3 and
this can become 1. So, both are fine ok. So, these are the only options you have in this
particular network. Now, what you will understand is in this case the resources were in
such a way that I can actually you know swap the resources from activity A to B to C
and D, I can share my resources on all the activities ok. And, what sort of examples can
you see in construction site for these cases?
Now, let us take a highway widening a project this is example we have seen earlier for
LOB or and so on. So, in this example what happens is this is a highway widening
project. There is an existing carriageway here and this is primarily the carriageway
which is proposed for widening and I have my embankment plan to be here and these are
all the levels with which I have to fill my layers on the highways ok.
And, the order of activities is excavation, then embankment, subgrade. You will see the
same thing here. The embankment is here, sub grade, granular sub base, wet mix
macadam 1 and 2, then dry bituminous macadam and then bituminous concrete and then
the highway is actually ready.
If you see the resources, now one thing the resource problems will be very common in
repetitive projects because I am going to use the same set of activities. Now, this set of
activities will be there for a stretch of 300 kilometers or so on. So, you may be having
repeating the resources on all the places on all the stretches. So, this combination itself is
repeating ok.
Now, if you see at the individual activities also. I have so many resources which are
repeating I have only dumped in the common resources ok. I need excavator and all
these are serial activities which implies sequential in execution. Excavator and grader –
excavator is again required it for embankment; grader is required for embankments
upgrade and granular sub base. Dumper, I need almost for all activities except my
excavation and roller I need for three activities embankments sub grades and granular
sub-base. Paver I need for the last three activities like wet mix macadam, dry bituminous
macadam and bituminous concrete.
So, if you see here there are common resources which you need. So, if you are planning
if the number of resources are limited then you may have to execute one only one path or
you may wait have to wait for the previous activity to be done and then only you have to
go in for the next activity even in the sequential and only when full stretch is done then
only you can go to the next stretch and so on. So, you have to have a so much of delays
may happen when you are having common resource shares ok.
Now, let us see a real example as to how this you are actually working in construction
sites ok.
So, now this is an example, this is also a known example. I am bringing in all known
examples to you and let us see the resource requirement here. This example I have
brought for 3 tower units. I am going to show only 2 tower units as the earlier. Duration
for all these activities are listed here. I have used same duration because that is the
assumption we generally make in LOBs I have brought in the same duration and IPA for
all these activities are listed here and if you see here the resource required I have
different skilled workers, but I have same type of an unskilled worker 15 helper, 10
helper and 10 helper I have the same combinations in the three activities.
So, I am going to show you what is the option I have only for the unskilled workers right
now and how the duration is getting changed or modified. So, the skilled worker we will
see a little later.
Now, let us see only for the helpers. This is a resource aggregation chart I have right now
and I have placed this for activity as well as I showed you for the aggregation also ok.
Case 1: So, what happens let us assume I have almost the maximum number of crews I
want I have with me. There is no problem, no constraints on resources. So, 3 towers have
to be done. So, all the 3 towers can be done in parallel. So, what happens all the 3 brick
work is done in parallel, all the 3 joinery are done in parallel, all the 3 flooring are done
in parallel, all the 3 painting tasks are done in parallel because I have enough number of
my skilled crews and workers ok.
So, if you see as I told you we are only going to see on helpers, if you see here the
fluctuation of the helpers which is shown here. This is on helpers total 45, 45, 45 then 30,
30, 30. Then no workers are required for painting job and the duration for this entire
project has consumed to be 330 for all the three towers until painting ok.
If you see so, for example, till level 2 at least if you want to finish, then you are stopping
at actually 255 days I am finishing only for the three activities till level 2 ok. So, to have
a fair idea on where we are. So, 45 workers I need at least for few days of my whole
project.
Now, let us assume I have very scarce resources I cannot go with 45 and the maximum I
can do is 35 or something. Then what you have to do? You have to stagger the execution
of the whole work. So, what I am going to do I am going to follow the little of the
conventional repetition works how we do after one level of brickwork, the second level
of brick work is done then the third level third tower goes. So, after joinery the first
tower, second tower, third tower so, after the flooring first tower second tower and so on.
So, this is a bar chart representation I have shown you for the same ok.
Now, the same only for the helpers I am see showing it here. This is the total on all the
workers whom I have. If you see here the maximum limit I have is only till 35 ok. I have
consumed only till 35. Now, what about the duration? So, until flooring for the three
levels if you want to see, my duration for the entire project is coming out to be 435.
Now, in the earlier case when there was no problem or constraint on resources I finished
until level 3, till tower 3, flooring activity was 255 ok, here I have stretched till 435.
Now, let us see still I have more constraints I am going to reduce it to 25 workers only.
What will happen?
So, what happens, only in this place I need more than 20 workers or so. So, this
brickwork cannot go in parallel at this particular time. So, I am just shifting the
brickwork compared to the last place to this place so that I maintaining the maximum of
25 a limit ok. Now, what happens here I have not distributed the resources in the
previous places and adjusted with 5, 5 and so on. So, the leveling really I have not done I
have only shifted the work completely ok.
So, what happens here when I have stretched a little bit so, now, until till your flooring
activity for all the 3 will go till 510 I have cropped only till joinery, if you add flooring
painting and so on. So, it will actually stretch till 510 until flooring for the 3 towers
which you are planning for ok. Now, this is what is the whole exercise all about.
Now, let us see further. So, what happens here is the logic you can actually you know
this logic alone has to be maintained and you can actually move all the towers in parallel
also if you have more crews. So, primarily the number of resources what is driving your
duration which is what is the resource scheduling all about. So, starting from somewhere
around 330 days of duration ok, we are actually in to then 435 and then till 400 and 510
for this case.
So, we are having a large variation in the duration because of the number of helpers you
have in the whole project ok. I am not even talking about the skilled workers now. So,
now, what do you do? So, there is unique so, primarily this is not the hard logic it is only
the resource relationships which you have to represent in the network.
So, how do you represent the resource relationship?
Now, let us talk about the RDM diagrams. So, RDM is nothing, but Relationship
Diagramming Methods. So, in any network diagram the logic dependencies is primarily
driving the whole sequence. Now sometimes the logic dependence you cannot alter the
logic dependence, but the resource dependence can be altered based on the availability of
resource at any point of time ok. For example, in the same apartment exercise you can
have all the 3 towers going on in parallel or little staggered or too much of staggered
based on the number of resources.
But, as such if you want to draw a diagram and represent this how do you draw the
diagram? Do you put a link or you show it as parallel? So, how do you really do ok. So,
RDM is the only one way to show your resource dependency on a project ok. The reason
why code on RDM can be associated with their constraints: so, primarily choice for the
reason why R either I go with the P for physical or R for a resource. So, primarily you
have right on the network whether it is P then the P the violation on FS is never altered
and if you are writing an R in the relationship based on the number of resources that you
have the R can be made parallel or it can go sequential or it can go even pushed and then
started. So, that is what happens here.
Now, the same example now I am going to show and on the skilled workers case ok. So,
in the same example what happens let us assume because I am not worried about my
helper I know I can easily get my 45 helpers or 50 helpers or how many ever I want in
the whole project. I am only worried about the skilled workers because they drive the
whole activities ok.
So, I have just marked only on the skilled workers here. So, maximum availability per
day 15, 15, 15 and 20 that is what I have written here. And, so, what happens here? So,
can you start all the activities in parallel if you look at what is the requirement of these
workers I need 10, 14 for joinery works, 12 and so on.
So, primarily you can go with one activity, 1 tower at a time, but you cannot do 2 towers
in parallel and what if since I have little more than 10 mason and little more than the 14
carpenters and so on, so, can a little more push some of the activities in the tower 2? Yes,
it is possible. So, what I have done, I have used combinations 1, 2, 1 so on. So,
accordingly you can do this diagram. This diagram is primarily called an RDM diagram
and you can still do calculations on what is your critical path and so on. So, in this case
with this combination this happened the tower 3 happened to be the critical path ok.
Now, if you see here so, primarily what you have to do on the arrow mark put a slash and
then from there you can write what is that relationship all about ok. If you see here I have
written here physical here because after my superstructure only I have to do my brick
work in tower 2 and I have to start my brick work in tower 3.
So, there is a physical relationship also here. This red is primarily the critical path. And
after the brickwork in tower 1 and brickwork in tower 2 masonry is a resource which is
governing there ok. So, if I have more mason I can do both the activities in parallel. If I
have a few mason then accordingly this turns to be sequential in nature; accordingly the
same you have for brickwork in tower 2 and tower 3 ok.
Now, same thing for your joinery works ok. So, I have my carpenter: so, carpenters is a
resource for my brickwork and so on. So, what happens after my after my joinery work
is done then these carpenters will be playing a major role for my resources ok. Same way
for flooring I have tile layers, painting I have painters; so, more painters so, this piece is
like a flexible relationship I have and this is a representation on an RDM network ok.
I am not showing the calculations here. So, when I change the combinations a critical
path also changes and so on. So, that is primarily on this RDM.
Now, let us see the resource constrained schedules ok. So, far we have seen schedules in
which duration of the entire project was not altered even if it was time constraints or it
was a resource constraint, resource which was constraints ok. So, you try to
accommodate within that only.
So, now, the next case is this suppose if we have very scarce resources as a result of the
very few resource the duration of each activity has to be stretched little bit and
obviously, it slips off beyond the planned schedule of the whole project ok. Those
particular cases we call it as a resource constrained project scheduling problem or
resource restraint scheduling. In some cases it may not be possible to schedule the
activities so, that the maximum number of resources does not exceed the available
resources. So, in these cases it may be necessary to extend the overall duration ok.
Now, let us take the same case. So, the same hypothetical example which I showed you
earlier. So, this is primarily it takes 11 days of duration ok. So, this is resources, number
of resources and this is time in days ok. So, maximum I have taken 11 days and my
resource was primarily maximum I have taken consumed is 5 ok.
Suppose, if I want to make it as 4, what happens? Still it is possible within with very
little floors we can still make it possible and I could still adjust to 4. Till then you still
call this problem as a resource leveling problem only because you have not extended the
project duration any on any case. So, this we call still as a resource leveling problem ok.
The minute you start having very little resources maybe let us assume 3, what happens
my duration of A or D gets stretched a little bit and what happens it may even extend my
critical path durations ok. Now, it has gone till 12 and the number of resource consumed
on the whole project was 3 here. So, these type of problems we call it as a resource
constraint project scheduling problem.
Now, what are all the approaches to RCPSP? There are so many approaches available. In
broadly I have two methods for RCPSP; one is optimization approach, the other one is a
heuristic approach. So, optimization approaches what does it say? It seeks the best
solutions but, are very far from limited in their ability to handle complex situations and
large problems. In heuristic approaches, they employ a rules of thumb that have been
found to work reasonably well in similar situations.
So, a lot of heuristics are there; one is simple heuristics, another one is complex
heuristics. Simple heuristics; so many rules are there in the form of rules as. ASAP as
soon as possible; so start finishing up the activities as soon as possible. Then, yeah
ALAP; so, primarily it is as late as possible; then most resources first; minimum slack
first; most critical followers which implies whatever activities which have many
successes, then those activities should be done first on the resources most successors,
then combinations of any of the above also is possible ok.
Complex heuristics will simulated annealing and Tabu search, genetic algorithms and so
on. So, so many methods are available on RCPSP modes ok, but this has to be know
dealt in a different way in order to work it out on problems.
Now, let us see on modes. So, we are actually you know there is another simplistic way
of working out on the RCPSP problems which is their existing for last 30 - 40 years and
so on which we call it as a mode. So, what do you mean by mode. Mode is the possible
way of executing any activity. So, I have two combinations here; one is called the single
mode RCPSP problem and the other one is called multi mode RCPSP problem ok;
accordingly you have SRCPSP or MRCPSP ok.
So, for any activity there is only one way of executing the activity, then we call single
mode. Completely for the entire project generally people work it out on only one way of
executions. Suppose that the resources are constraining too much why to look at those
resources in a different combination maybe there are other resources which are not too
much of a constraints is it possible to think of getting that resources and then doing the
scheduling in a different way, that is primarily called the multi mode options ok.
I will show you with an example. Now, this is an example hypothetical example just to
show you on how to do the whole exercise ok.
Now, this is the example. I have a list of activities here. These are the IPAs for the
activities. Resource requirement in a man-days is given like. This I have a crew size. The
3 combination crew size; primarily the 3 mode options I have; minimum normal or
maximum; I am taking it as my 3 modes right now and the man-days are given here.
Now, according to the same problem which I have here what can I do? I can calculate the
duration for the whole exercise ok.
You know what is the crew, what is a work content, what is a crew size if you know so,
based on this you can really calculate the duration for all these activities on the minimum
normal and maximum crew size. Now, I have got the duration for all my activities ok.
Now, let us choose some combinations now. Now, what am I going to do if you see this
example it looks like A, C, D is on one path and activity B and E is on the other path.
That is how the relationship goes through. After the A, I have C, then I have D, so, A, C,
D is on one path and B and A is on the other path. Can you think of critical path and
other combinations here?
You cannot think of anything here because it is like you can choose any options you
want ok. Whenever you are working on mode there is nothing called critical path
calculations in when you are working on mode calculations.
So, I am going to choose a crew size on 4 because I had a common number in all the
three activities. So, I have taken up that as a 4 and what happens I am going to choose
those duration so, which implies 3, 8, and 5. So, when I am choosing a resource
combination on 4 along one path. So, 4 on A, C, D my duration for the A, C, D will be 3,
8 and 5 which is nothing, but 816 ok. Now, when I am choosing so, same way I am
going to choose a common resource on B and E. So, I am going to choose 6 and 6. So,
accordingly what happens my duration is 10 and 6 which is again 16 here ok.
Luckily, I had a very good levelled path here with the duration of 16 and my resource
combination is actually 10 here ok. What happens, the same options I have taken here
and so, these are my activities I have done and this is my complete till 10 is my resource,
ideal resource combination I have got and my duration have come until 16.
Suppose, if you want to use more resources and you wanted to reduce the project
duration or you wanted to use more resources or you want to reduce a project duration in
any of these combinations what happens is you may have to know you may have to have
a unleveled combinations will be there, but you will have so many options in order to
evaluations.
So, in real world practice what happens, just I am taking on one resource here and I have
shown how to do, luckily I got in the first trial itself. You may have to do so many trials
to see what is options or evaluate all the possible three options for all the 5 activities and
then see what is the best combination for you ok. This will turn out to be the best ideal fit
for you in any case.
Now, let us take this another example in order to show the complexity with many
resources coming in ok. This is a steal truss bridge ok. Primarily what happens, this is a
simple steel truss. So, I have a steel truss here, this is my deck slab, these are my
bearings I have a well foundations here, this is a pier cap and this is a pier. So, primarily
I am going to look at pier and the pier cap activities only just concreting ok.
So, I have taken I am going to look at the activities only on the piers and pier cap, only
the concreting activity alone equipment at site. So, these are the equipments I have a
batching plant, 3 excavators, 8 dump trucks, 1 concrete pump, 4 transit mixers, 1 boom
pump and 1 mobile crane I have in the site ok. Resource on manpower I have 30 bar
benders, 12 20 carpenters, 30 mason first grade, 15 mason second grade and 8 pumpers
and 50 unskilled workers I have in the particular case ok.

Now, I have four activities for these two four sub-steps or sub-activities for the main
component one is pier and other one is pier cap ok. So, rebar for pier, shuttering for pier,
concreting of pier, de-shuttering of the pier ok; again rebar for pier cap, shuttering for
pier cap, concreting for pier cap and de-shuttering for the pier cap ok. Now, I have two
combinations here; one is mode 1 other one is mode 2 and for the different resource
combinations I am getting a different duration.
So, this data you can get from past projects from the experts, when you go to the expert
they will actually have an option second alternate option of evaluating the same activity
when the first option does not work out. So, this is primarily the preferred option and this
is an alternate option data which we have on the site ok.
Now, what you have to do, you may have to evaluate all the options to in order to
evaluate what is the best cut scenario. One option is to consider all the activities in the
first mode, other option is consider all the activities in the second mode and maybe this
activity C and H there was no other option, second option was there so, I have taken that.
So, accordingly choose combinations on activity A first mode, activity B second mode
so, activity C primarily only one mode is there, now D you choose one time first mode,
one time second mode.
So, you will have n number of combinations available and you may have to explore all of
them if you are not going with the simple heuristic methods there ok.
Now, what happens in certain cases? In some cases what happens is even the resource
types also vary what we call it as a generalized multi-mode resource combinations ok. If
you see this example that this is this is also a simple activity example in the housing
project ok.
So, I have column marking, column reinforcement, shuttering concreting, de-shuttering,

staircase shuttering, slab shuttering, reinforcement, conduiting and so on. If you see here
I wanted to highlight you only on the mode 2 only for these particular combinations ok.
Column reinforcement I can do with 6 bar benders, it takes 12 days. I can also have a
combination of 6 bar benders and 1 crane duration of 8 ok. Shuttering; 8 carpenters it
takes 6 days, I can still do the same with 8 carpenters and 1 crane with a duration of 4
days. Reinforcement; 23 bar benders or 7 day combinations or 23 bar benders does an 1
crane on 5 day combinations ok.
So, now, what happens here, in the first case any activity can take any resource. Problem
was too simple. So, resource scheduling, scarce resource itself is a real complex problem
ok, little simplistic case is any activity I can dump in any resource then it was still an
easier case. For example, in these activities, all the activities whatever are there they are
having a different type of skilled worker and so on, here you cannot really shuffle the
resources ok. So, then the problem was little complex.
The next case came in wherein I had multiple options to execute and I had a constraint in
my resources and I have multiple options to execute still you can look at options in order
to see what is the minimum duration and with my constraints on my resources what is
my best fit operation available option for me you can go ahead with ok.
Next option is even in the same modes I may have different types of resource
combinations also for my modes. For example, I can do painting by using a brush
manually or I can use and spraying for mechanism for spraying and painting in the whole
case ok.
Now, I have two options. So, based on the type of resource what happens if I have many
constraints on my painting resource on my painter I can still go with them you know; I
can still avoid my manual method of painting and I can go with the other resource. So,
what happens resource constraints is really solved, but another case can crop up which
implies the cost part.
So, what is the cost if I use another crane in the whole setup only for doing few
activities? Duration can go less, I am having actually no constraints on my resource, I am
able to finish fast, but my cost can go up ok. So, primarily you are scheduling in this case
you have to look at the combination of time, resource and cast together then only you
will have a very success in projects ok.
So, with this I am stopping for today’s class. So, today’s class on resource schedules. I
showed you a few trial and error methods only because the scientific methods are little
complex in nature. So, I have not covered and interested people can go through research
papers when you are interested ok.
Bye.
Lecture – 07
Information-driven Scheduling
We will see about Information-driven Scheduling. So, what do you mean by

information-driven scheduling? So, first let us discuss on where this generally happens.
So, this primarily happens in the design phase.
So, we will see a little introduction to the design phase and how far it is a little different
from the construction phase and so on. And, then what does this information all about so,
that also we will discuss in little detail with some examples. And, then the very famous
tool for managing information-driven scheduling is dependency structure matrix. So, we
will see how to represent it and one step in the DSM is formation of an activity DSM that
also we will see in today’s class.
So, now let us start with project execution strategy. So, conventionally if you see what
happen is, so, design is generally done and once the design has done, approvals on the
design are taken then only the good for construction drawings are generated and then the
construction really starts that is what is a traditional method of executing any project ok.
What happens as a result of this, you have to wait for a lot of time for approvals and so
on. So, it takes so much time for a real project to be completed ok.
When people are really interested in completing the projects faster and so on, so, there
came a methodology where in design and construction can overlap. There can be so
many ways of overlapping and one way of accelerated (Refer Time: 01:47) construction
technique is what we say is a fast track method; it is one method in that group. So, what
happens in fast track, we do not really squeeze the design phase or the construction
phase? We try to overlap as much as possible wherever we can overlap we try to do it
and as a result there is a little time savings in the whole project and that is where the fast
track came in came into existence.
And, since even then this was not at all know enough for the owners or the contractor.
So, what happened is later on came in came in a way to even still squeeze the total time
for completion. So, construction we would not be able to squeeze too much because of
the hard logic constraints only after the foundation you can do columns then you only
can do beams or slabs. So, that hard logic you cannot violate or you cannot no go away
with that hard logic unless you are thinking of prefab structures or modular units and
then assembling it in the site even then the hard logic has to be maintained.
So, construction phase as such you would not be able to squeeze, but what happens in
design phase? Design phase is all about information between the people. For example,
when I want to release my foundation design, I have a small structure. I have a
foundation design to be done. I may have the superstructure design in several forms I
have ok.
Suppose, if you want to do foundation design first because you have to give it to the
construction site for them to start the construction, what can happen? With your earlier
assumptions you can always assume the loads on the structure and what are the live load,
dead loads and seismic loads, mean loads whatever you want you can assume you can
still release the design for foundation earlier to the sites ok. And, then later on with the
real design happening in the substructure and other parts of the whole structure then you
can recheck your values and then you can modify here and there small segments, but as
such the excavation, foundation can all happen in the meantime ok. That process we
generally call it as a concurrent engineering.
Concurrent engineering is the engineering phase or the design phase is generally

squeezed by putting all the teams to work in parallel. This is a very common
phenomenon right now in many of the projects which are happening today because of the
speed with which we have to finish the projects. Concurrent construction is again
squeezing within the construction phase through I said alternate techniques are available
with which you can squeeze. So, as such the complete construction design to
construction phase can be really cut down a lot ok. So, these are all the strategies.
Now, why I brought this picture here is to show you that there is a hard logic really
driven in construction phase, but in design phase is primarily the information which is
driving through. So, as much as squeezing is possible in design phase compared to the
construction phase for that only I brought this picture right now.
Now, let us move on into what is this design management all about and what is it
actually it is to people. I am not talking about the design phase of a construction project I
am talking about the generic design management ok, it can be a product design, it can be
a process design, it can be an aerospace design. So, I am talking about in general right
now.
So, it is a business side of design ok. What does this mean? So, design managers should
have to speak the language of business in order to know sell their design they have to do
little on the business issues on what will attract the customer, what will attract the
stakeholder. So, he has to do that language of business and also he has to do the language
of design as well so, which is a very challenging task. You will see later. There are lot of
techniques available even for the synthesis portion of designs.
In the last 15 years, 50 years sorry, design was mainly with managing the aesthetic issues
on design which you always know. Whenever we talk about design, generally no time
frame is really given for a specific design solutions and analysis because it was generally
aesthetic appearance and construction was dominating the whole show that is because of
the large time it takes for completing it and also for the large cast which goes into the
construction phase, that happened in construction also it happens in other domains as
well.
So, it has been formalized the design management has been recognized as an individual
discipline only since twentieth century or even later than twentieth century and so on.
And, what happened after then also there is a slight shift in the focus and now managing
design is all about ensuring the ability to innovate how far I can build a new structure,
how far my construction will be known everywhere. So, iconic structures or a new way
of doing the whole project everybody wants to do something new compared to what
others have done. So, now there, your focus again has shifted a large from the last 10 –
15 years or so on ok.
So, now, what is this design problem all about? So, it is not a regular design as such, you
have to have several other issues behind. For example, it talks about science many of the
ideas comes from natural things it is primarily from the science. From the birds we who
can fly from one place to another the idea on aero plane came. So, many of these
concepts are really coming from natural sources and from the natural science which we
have ok.
Then, next is art. How can you present your artistic view into an idea generation, idea
concept, a conceptual issue? So, so from there, art also is part of the whole design phase.
The next is engineering. So, if you want to build everything if you want to put in your
imaginary world which is your artistic world and your natural world into in, then you
need to have little more on artificial things coming in it can be in form of steels, even in
concrete. So, what is the right combination for me to execute or put in together in place,
so that it serves a real purpose on what I wanted to achieve ok. For example, the shape of
a lotus, you have a lotus temple. So, like this so many things have happened, but you do
not get the same feel of a lotus, but there are primarily the artificial things which are put
in place and which gave you that feeling of a structure.
So, design primarily you have to have all three characteristic features put together I just
brought on small video art view here to show you on this. So, what is a design?
Primarily, it is to formulate and execute a plan to satisfy a need ok. So, I should have a
proper idea on what to do. So, the artistic view cannot be really an imaginary part, you
have to have a little synthesis in your mind and you should also be able to execute a plan
and that should be a need which is driving. So, primarily there is motivation or list of
needs which is driving me to generate the design ok. I need something new which
nobody has done earlier. So, the even the innovation is also drived or driven through the
user needs or client needs.
So, this design is generally born when the customers motivation is stated in terms of
needs ok. So, it can be in so many forms in so many ways. So, let us see this, to design is
to reduce uncertainty through a decision making process as I have shown you in the as I
have dealt earlier examples also even in the WBS example on developing a laboratory I
have shown you several steps on which so many decision making are taken taking place.
Same way even in the design process you are actually reducing so many uncertainties
through variety of decision-making process and in design it is like you are actually
preceding the entire construction.
So, too many assumptions you have to make in because you are entering a construction
world is unknown to you. You do not know who the contractor will be, you do not know
what equipments are going to bring in, you do not know what technology they are going
to bring in you do not know what form of equipments or methods we are going to use in
the whole project, but you have to assume for a variety of people who can execute the
whole project. So, there are a lot of uncertainties, but the real challenge is how much
ever uncertainty you can reduce at same time there is a decision making at each and
every stage.
So, with each decision, the designer must choose what he considers to be the best from a
set of n possible paths. For example, if I have to do a foundation. There can be so many
types of foundation, starting from isolated foundation, your raft foundation, pile
foundation like this I have n number of methods ok. So, which method I have to choose.
So, there are several paths for each and every step and you may have to cut down the
path or at least try to make it minimal as and when you go on with the information that is
pouring into the project. So, that is primarily a challenge you have on a designer ok.
So, next; so, as I told you there are two phases in design. There are so many tools
available for these two: one is synthesis. Synthesis we generally used as a solution to
arrive at your best or a feasible design solutions and analysis methods are also available
for example, for change management, for rework, for iteration or for scheduling
techniques. So, there are several methods available, since this course is primarily on
scheduling we are going to see the aspect in analysis only and synthesis I am not
covering here; analysis also only one method we are going to see in detail ok.
Now, this is also a very famous slide which you will have which you would have come
across earlier. So, there are so many domains available in a project especially when you
are doing a design ok. When I said set of n possible paths and n ways to do, so, what is
this all about?

So, I have a customer domain ok. Customer domain or a client domain both are really
different. Customer domain can be superseding your client demand. Client demands may
be most often the basic demands of a project. For example, I mean auditorium, or I need
a house ok. That can be real, basic and mandated requirements can all come from a client
domain, that is not what we are talking about. We are talking about the next superseded
layer over the client domain which is primarily called customer domain ok. I should have
a house which can rotate among on it is axis or something like that. So, they all can be
the added features which you wanted to have along with the basic demands of a house
the functions of a house.
And, then the next is functional domain. Suppose, if I want to create a structure which
will serve as an auditorium then it should have all the acoustic features, lighting issues
and so on which really know that I could feel that it I am sitting in auditorium ok. So,
there can be W needs on my functions and there can be Y responses. If I want to do all
these performances what all I can do? I can have acoustic issues for example, I can have
instruments, or I can have a different form of door setup or I can have changed my all
wall patterns differently or I can have something which can absorb all the sound and
other noises. So, there can be Y responses for all my W needs. So, for each need I may
have so many responses.
And, then what will the customer really want if I choose among the Y responses maybe
one or two he may be happy then you can only choose those is a responses that is
primarily the satisfaction level. Some are basic satisfaction levels and some are like
hidden satisfaction. Suppose, if you attack all those responses then the customer will be
over happy and over enjoyable.
So, there are now three domains in a functional region itself – the W needs Y responses
on Z satisfactions. So, primarily you have to map between them and do a decision
making, that is what mentally they do and then arrive at a solution space which need not
be a single solution space. It can be a set of solution space as a solution we have driven
which we can really execute. We have to also think about how to do the same solution
spaces, is it feasible, can I have the resource to do, will I have the budget executes all
these you have to evaluate. So, primarily you land up in one or two solution spaces only
which is the physical domain, the real execution part of your whole project ok.
So, that is what is the spaces and definitions. So, which primarily you deal with all these
in your synthesis solutions ok, but since we are not covering them so, I am just moving
ahead.
So, what is this design specifications and this is a term which you heard you would have
read in many of the construction starting classes as to the estimation or so on it starts
from design specs ok. So, what is this design specs and what are the related terminology
with the design specs because we are really seeing into the design phase.
So, explicit requirements are primarily defined and collected. Explicit requirements are
the written stated which are very clearly written that I need a house which should be very
secure and so on. So, all these are primarily called explicit requirements. These
requirements can be analyzed, they can be refined, they can be extended whatever you
want you can still do with them ok. There are certain related terminologies with these
explicit requirements which we call demands and wish or wants and needs. There are so
many colloquial terms which are available in literature, but I have brought few
terminologies here.
Explicit requirements are those requirements given by the client which are really
explicitly stated in the in the in the database ok. Implicit requirements are those
requirements which you should be able to grasp as an when the client is explaining about
the project, ok, these are all he may know wish to keep it in the project. So, you may
have to grasp the series of discussions or interviews with the client. You should be able
to rightly capture what are all the hidden I would not say they are the hidden
requirements, but they are not explicitly written, but they are part of the whole
mandatory requirements ok.
The next is obvious requirements which may also come from user demands ok. So,
suppose if I am living in some place where there is so much of rainfall and dampness and
so on, then I may have to raise my floor level little high, so that the water or something is
not coming into my house or touching my house and always it is damp. So, these are all
user demands and which are obvious requirements which may not come explicit, but it
has to be attacked by the by the designers.
The next is hidden requirements. Hidden requirements are only for attractiveness ok. If
somebody wants to buy your house or you want to sell it to someone so, what is the real
attractive feature you want to keep in your house so that it is attracted by an n number of
customers that also you may have to study. Now, when I am saying about talking about
the customers then there are two types of customers do you want to talk about all
anonymous customers then you may get so many responses.
For example, there was one apartment renovation in our campus, and we wanted to know
what the real users wanted in the same apartment. So, we did a survey; it was a sample
there was one apartment block, 12 houses. We did a survey and we wanted to know what
all these occupants wanted when the houses are to be renovated only the interior
renovation was planned ok. So, these 12 occupants they gave almost like 96 or 100
responses and which were like two distinct from each other rarely one or two were
common requirements from all the occupants otherwise many of the occupant occupants
gave a different set of user requirements.
If you wanted to really implement all these then what happens to the budget of the whole
renovation it may shoot up really high ok, but these are all specific customers who may
stay in the in the same apartment when the houses are renovated. So, these are all
specific customers. Now, if you want to take out real other sample survey wherein some
other block of people who wanted to have a renovated house so, what can be their
requirements? So, these are all anonymous customers who have not stayed in the same
block, but who can give a different set of a user requirements.
So, whom do you want to really attack, is it anonymous customers or specific customers?
Now, the challenge is now in front. Now, again what we did was we did a priority on all
these requirements and we thought of doing at least 10 of these requirements were
prioritized and then we went again and roamed around in the entire block and seeing
whether the other requirements was also part of the customer. So, some occupants they
change their mindset and they said yeah, these are also part of our wish list and they
change something, something got modified and then we ranked a 10 requirements to be
done and we did a survey on how these and 10 requirements could be accomplished.
So, like this it is really a challenging task when you want to really attract with the hidden
requirements. So, most of the designers they generally do not attempt to do, but primarily
it is also part of the whole project if you wanted to really have a nice project to be done
ok. Then these requirements can also come in two forms – one is quantitative form, other
is qualitative form. Quantitative form is really easy to do, but qualitative form is really
very difficult to do. I want nice air circulation. The nice can be in any extent you wanted
to push or pull so, which is very difficult for anybody to know to handle all these
qualitative requirements on the customer.
So, these are the general issues on design specs which I which you should understand
when you are working on the design management or information-driven schedules ok.

So, the next is stages in design. So, obviously, you need to know what are the different
stages and what is this design process all about. Design phase is generally termed as an
evolutionary process. You start with the customer requirements or demands and at the
end of your whole design phase you are just generating some 500 – 600 drawings and
along with a set of specifications. So, what happens in between that many of the people
are really not very clear.
So, the primarily there are three stages in design phase: one is a conceptual, next is a
schematic and the third one is a detailed design phase. Conceptual design phase – It
results in specification of a principal solution or a concept. It is slightly ahead of your
feasibility stage of the whole project and as soon as a feasibility is done conceptual
design stage starts and de-marketing these three stages are really very difficult. It is like a
very hair line mark I will say on the three stages. Generally, if you look at a project
primarily it can be a J or a layout diagram or something or a line sketch which we
normally say that set off an information should have been finalized when a conceptual
design has been completed. So, the principal solution is primarily achieved here in the
conceptual design.
The next is schematic design. So, schematic design results in the specifications of the
layout; ok, where are the arrangements, where should I put my boxes, where should my
apartments come and how should the three layouts look like, what is the orientation, and
what is the levels, how many floors are there, how many rooms are here. So, primarily it
talks about little more specifications on the layout and little more details have come.
Primarily, you would come to know what are all the components and how a structure
will look like, partially you can have a visual imagination and idea, then you call it a
schematic design ok.
Detailed design is primarily you have to work out on all your design calculations
analysis and so on. You work out on loads, you do the design. So, primarily this is a real
detailed phase and it takes more time and which arrangement, forms, dimensions, what
type of concrete I am going to use, mix designs everything are completely done and then
they are all laid down in the form of drawings aspects. Only in this phase you really get
dimensions and the properties or layout detailing of the spacing, reinforcement spacing,
type of reinforcements all these generally comes in the detailed design phase.
These are three stages in design in ok.
Now, next is the design thinking. Generally, what happens you have something called a
permanent structure design and you have something called a permanent structure
construction. In permanent structure design what happens is whatever drawing space you
do for the real structure that is what we call it as a permanent structure design. In the
permanent structure construction, you really think of foundation this that for the real
construction, but what happens there is some intermediate steps here which is called the
temporary structure design.
All your temporary structures like scaffold, centering, shuttering, false ceilings, sorry not
false ceiling you and all your support structures like scaffolds, shuttering, centering and
all these are primarily your temporary structure design and those erection installation of
those structures are the temporary structure construction. These generally are ignored in
some projects, only recently these are all part of the whole construction situations.
Now, what happens is if you want to really do an unique design or design which will not
fail through whatever methods and means you are going to adopt in the construction
stage, then you may have to think for what can be the temporary structure design, the
temporary structure with which they are going to construct it. Is it an aluminum form
work or is it an steel form work or is it an plywood or timber form work or so, those all
you have to think about and then what about my permanent structure construction and
then you may have to have an idea on that and revisit your design and correct all your
designs accordingly. That is primarily a challenge which you have in many of the
projects.
And, some projects where some project managers they really do revisit on design
thinking for constructability. Otherwise what happens is the minute you start your
construction every time you may have to re-do some parts or redesign some parts and
you waste a lot of time and cast as a result of the redesign and the works ok. So, what
happens in our project and what is this information. So, we are just traveling into the
information management as such.
So, let us take a small project. This is an underground metro construction project. If you
see here there are like 20 teams, this is even more. This is on a planned schedule on a
project which takes 3 years what is planned so, which is given in quarter. So, year 1, year
2, year 3 are the three years timeframe which we worked on the whole project and is
given in quarters. If you see the entire 20 teams are almost traveling through the
complete all the 3 years of project.
Now, the teams were project lead team, planning team, billing team, contracts team,
design coordination team, interface management team, form work team, quality
assurance control team, P and M team, environmental health safety team, survey team,
architecture design team, structural design team and so on. So, primarily it is a series of
design team, construction team, installation teams, survey team all are traveling together
in a metro. In a metro construction you do not do the design first and then you go. You
go parallelly with the design and construction.
Now, what am I going to say here? So, even in construction you have so many people
you may have to coordinate, and you may have to work together. So, then what is a big
deal in the design phase? In design phase what happens is you may have to work more
with the information phase which is what is challenging. Let us see that right now.
Now, what is a design phase? Again, I am just moving on into an example. This is an
offshore facility example. The same project has a different view from different people.
There were 8 teams in the project like architecture, HVAC, electrical instrumentation,
piping, mechanical, structure, health safety environment process. So, like this there were
8 teams working an offshore facility project and what you would see here is all the 8
people had a different view all together on the whole project. Now, bringing together
their views and doing the sequencing on the whole project and getting the information in
proper order and working out on the project is always a challenge for any project
manager ok.
Now, this is on the team level. The last slide I showed you on all teams almost work
together, this is another example again showing that different teams have different views
on the entire project and which is like not even know matching with any of them ok.
Now, let us see little more in depth what really is happening. This is a highway design
project again ok.
So, here if you see what really happens is each and every day almost like so many
information is traveling across different teams and across components and so on ok. So,
primarily we call this as a 3C or 3I problems. So, 3C is primarily your communication,
collaboration and cooperation. Some people also extend this to complexity, co-location
and so on. So, all these are primarily your 3C’s ok. We will come to the C’s later.
Now, let us talk about the 3I’s. The I’s are primarily information, interdependency and
iteration ok. Now, what is this information? So, see if you see here ok. So, this let us take
a good example. From the vendor I have a bore log details which is given to a
geotechnical team. So, this is required to geotechnical team between the vendor team and
the geotechnical team I have a bore log details as a common as a parameter exchange or
a parameter transfer between them ok. So, that is primarily the information I am talking
about an information can come in any form. In this case they are the parameters.
What is this parameters I will give you an example also. So, let us take the design of a
slab. What are all the parameters I will tell you. At the end of a calculation on design of a
slab what all information you get those are all called parameters or they are primarily the
information on the whole project. It can be depth of the slab, width, slab dimensions
primarily and the reinforcement data, cover to the reinforcements, spacing between the
reinforcement, mix design of concrete. So, all these can be part of your information or
parameters ok. So, the small calculations which you go which comes with the value they
all primarily called the parameters ok.
Now, what happens the next is interdependency. Interdependency is when these

parameters are really interlinked with two or more entities ok. For example, I need some
information from another person, other person also need some information from me.
Then we have to work together in order to exchange and share you want to do like this I
will make an assumption, I will do something then you start giving me information. So,
primarily that is set together or working together happens when you have an
interdependency ok. In the same example if my assumptions for executing
interdependency goes wrong, then I may have to revisit this and rework for a few
segment for a few stretch in length that is primarily called as an iteration. In the next one
or two slides I will explain what is a 3I’s ok.
Now, what happens is I may exchange for example, a person and I am one engineer and
there is another engineer who has asked some documents from me. I am sending through
email in excel sheet attaching an excel sheet and the same person has reworked on his
software part and he did something and he is exchanging the information back to me on
the revised values in a pdf form. So, these are the communication media I would talk
about.
So, what happens after then seeing this pdf doc you would not know what happened with
the earlier your excel file, why where are the changes are happened then what do you do?
You send an e-mail and for which he does not have the time to respond then he picks up
his phone, he answers your queries on the phone. So, what happens here is there are so
many communication medium which has happened ok. Already your information is also
traveling between the parties in. I am talking about only two people here and how many
rounds? There are several rounds which is happening different information are getting
exchanged here and some information are also traveling across several rounds and if
these are exchanged in different medium in each time then what happens is it is all like a
chaos real chaos happens in the design phase.
So, that is what is the challenge we have in the design phase and we need really solution
tools techniques which we with which we have to manage the entire design phase ok.
The available CPM, PDM or whatever you have you use in construction can handle only
the workflows or the hard logics they will fail if you want to use your information flows
ok.
So, now as I explained verbally I thought I will explain you with pictures also. So, the
independent and dependent CPM can easily handle this you know this known already. It
does not require any assumption for execution. The other one is semi independent which
we have used in PDM for explaining the PDM as well. Now, the next is the
interdependent and cycles or loops which is what we are going to see right now in
information modeling.
So, this interdependent is actually A is dependent on B and B is dependent on A. Cycles

are primarily more activities are coming in and there are so many loops inside then we
call it as a cycle or loop. This is not dealt before. So, this is a recent concept which
comes in only when the design phase has become more and more complex solutions
started coming in design phase. And, these require assumptions for executions. For
example, if you want to execute this should I start with A or should I start with B or
should I start with B, for whatever it is you may have to make an assumption ok.
So, for example, if I want to start A as my first activity, then I need to make an
assumption as if E has completed and this information should outcome. So, I am
assuming something and then starting it and then obviously, you may have to revisit
whether what you assumed is right or wrong or is it close to what you assumed and so on
ok. So, the other issues which I have not explained semi independent we have already
seen. There is something called multiple two-way linkages which you will see down the
line after the DSM is covered ok. So, I am skipping it for right now.
So, lessons learned from CPM or PDM. CPM can handle independent and dependent;
PDM can handle the semi independent or the overlaps to some extent. So, the what you
cannot handle in CPM or PDM is the interdependency or cycles or loops. Apart from
information that is something else which you I am talking about is interdependency. You
know what is interdependency; you know what the cycles or loops also. So, now, is this a
new aspect altogether and was it not dealt earlier. Same type of projects we have
executed 20 – 30 years also.
So, how was it done? Generally, what happens this is what happens. You make an
assumption here and then you assume that it has been executed and you avoid this
relationship at all and then you use your normal CPM or PDM for execution with. Like
this only we have executed the projects earlier. Now, what happens is now we have
solutions and techniques which can help you to model the interdependencies and loops,
we can also revisit and see how much is the rework which is happening as a result of all
these assumptions. So, now, we are able to handle the interdependencies, and which is a
very common phenomenon in design phase of a construction project ok.
Now, let us move into tools. So, so far we have I have introduced to you what is design
management. So, design management has so many uncertainties it is plagued with too
many uncertainties because the starting phase of a project and while construction
happens you know what are the teams, who has done the design, layout, structure
everything is really clear when starting your construction phase. But, design phase is
really with the imagination from the clients you have to really start and execute your
whole projects which is really challenging as such ok.
And, we have also seen what is the information all about and the 3I’s primarily. At least
you know right now what is information and interdependent. Iteration we will see at the
end of DSM, you will have a very clear idea on what is iteration. Communication also it
is really both information and communication when we are traveling together it really
complicates the design phase. I think that much you would have surely understood with
the last few slides ok.
Now, let us move on to the technique, with the DSM how are we going to handle
information-driven schedules. What is DSM? DSM is commonly known as design
structure matrix because application is predominantly in the design phase. So, earlier
when that tool came into existence when the technique came into existence it was named
as design structure matrix only. Then, people started working with the construction phase
also with for information sharing in construction phase for example; we have also
worked in underground metro construction. So, all these had information passing on
between the teams.
So, after some time the researchers started naming it as dependency structure matrix and
the D was replaced, design was changed to dependency structure. People also have called
it as problem solving matrix because it can handle information problems. Then design
precedence matrix, information dependency methods because it handles a lot of
information, incidents matrix. You may be working on matrix analysis and matrix
calculations only, so, people also call this as incidence matrix. It is also called the N
squared diagram because it has a set of N rows and elements, N rows and columns which
are happened to be one element. It can be an activity, it can be a parameter, it can be a
team whatever you want that is hence it was term does an N squared diagram. These are
all the other names for DSM.
Now, what is this DSM? It is a compact matrix representation of a system or a project.

This matrix contains a list of all constituent subsystems, activities whatever you want to
you can still keep it, but it handles only one element at a time and the corresponding
information exchange and dependency patterns between them. It is usually a square
matrix having equal number of rows and columns and same number of rows and
columns also.
History this is a technique which has come into existence only since 2000 and so on and
right now there are several people and several applications on DSM and down the line
DSM can be like superseding your CPM or PDM in future ok.
Now, let us take an example in order to see how to form this matrix or how to read the
matrix.
So, I am having an graphical representation on a project. You need not worry about what
is a starting activity and so on. It can be F, it can be C, can be A whatever you want to
start because the sequence of steps will help you to tell you what is the order of
executing the list of activities.
So, I am having a graphical representation. There are six activities here. First step is I am
just plotting a matrix. This is called a skeleton DSM. Skeleton DSM was nothing about
which has a rows and columns and no relationships are generally marked ok. I am just
using my alphabetical order A B C D E F. So, this but the A B C D generally shows my
sequence of execution. After A, I can do B, then I can do C, then I can do D, then E, next
F and the same list you will see along the column as well. So, since the number of rows
and columns is same it is generally called as an N squared diagram ok.
Next, I am going to plot my relationships. What is the logic with which I am going to
plot my relationships? So, this is arrow mark I have given here. So, you generally you
have to read from top to bottom which primarily gives information to. For example: so, I
am going to take B to D. So, this is B to D. So, so this relationship has. Let me use my
pen ok. So, this is B to D and B to D is here ok, then A to C. So, A to C is here. So, this
is A to C, then I have C to A. So, C to A is here. So, this is my C to A then I have E to B.
So, E to B is this. So, E to B is this; the next is E to D. So, this is here and this is E to D
then I have F to C. So, this is F to C ok; C to E is here, then this is C to E. I have A to F.
So, this is A to F. C to F. So, here C to F ok. Then I have E to F. So, E to F is here ok.
So, this is how I mark my relationships. In the website if you see it would have shown
another way which is like this both are same as long as you are comfortable in doing it
this way or that way. You should be sticking to the same representation till the end ok.
So, I am comfortable in putting in this way. So, I have used from top to bottom. So, C to
A this is read as E to B. So, this is F is giving information to C, this is like E is giving
information to D that is how I am going to read this ok; that is what I have written it here
as gives information 2 ok.
Now, what about the diagonal? Diagonal generally does not have any relationship, but if
you want to do scheduling then you can add durations of each activity along the diagonal
ok. For example: so, I am so, I can write here a duration of A, this is duration of B, this is
duration of C, duration of D, duration of E and duration of F. If I wanted to really
calculate my duration of this particular list of activities that we are going to see after few
classes ok.
Now, what happens here is I am having some X marks which are above the diagonal.
These are the highlighted ones I have shown here and all these primarily indicate to me
that I have to make assumptions. The minute if you are having all lower diagonal
triangular matrix, for example; now, below the diagonal I am having all my X marks
then it implies it is a general CPM or PERT. The minute I am having my
interdependency or my cycles or loops, then I may have X marks above the diagonal
which I cannot avoid and that primarily are highlighted like this in yellow which is
primarily indicates to me that I have to make assumptions for executing those activities
ok.
And, now what you should also understand is. It is fine. Try to minimize the X marks
above the diagonal. There are various steps for doing it in the next few slides you will
understand that. Suppose, if you are not able to avoid X marks above the diagonal at
least see to that you are pushing the X marks close to the diagonal, so that you are having
only one or two activities within the loop on rework. If you are keeping the X marks far
away from the diagonal many more activities will be falling in place into the X marks
and obviously, you may have been you may have to work in loops which you should
avoid ok.
Now, one example I would like to show you on the types of DSM. This was taken from a
paper which talks about positive and negative iteration ok. This is a very simple example
there is a beam and I may have to lay ducts for all my cables. So, I have just put the ducts
here and all these holes for the ducts are given in dimensions. X is primarily from the
starting of the beam to the first says start off the duct d is the diameter of the holes and e
is primarily the set distance between the center to center distance and h is the depth of
the beam ok.
Now, these are all given in mm. Now, what happens architect he gives his dimensions
the steel fabricator starts changing the dimensions. So, this happens ok. So, this is a live
example, like this so many information exchanges may happen and so many iterations
can also happen, this is what I want to call it as an iteration. So, this says there is a lot of
revisions which has happened with the teams as and when they started working on the
details of the calculations based on their acceptance and criteria levels ok. This is
primarily called as an iteration and you want to know as such what time this entire
process will take place that is what is the scheduling we are going to calculate in this
particular DSM topic.
Now, if you want to model this, what are the ways to model? I can really plot a team
DSM; I can really have a parameter DSM. So, how many teams I have so far listed here?
architect, steel fabricator, engineer, HVAC subcontract like this the list goes on ok. I can
put all these teams and I can plot relationships between this for the particular example
and with this I can really derive my time which takes for all these teams to interact ok. I
can also equally form a parameter DSM. Parameter DSM how many parameters I have? I
have four parameters in this case d, e, h and x just you can plot a DSM with the same and
this will become a parameter DSM.
Note down the relationship between these parameters and then you can get the scheduled
for the same project I can also have activity DSM, I can also have a deliverable DSM, I
can also have a component DSM. So, like this you may you can build so many types of
DSMs are possible, but whatever it is you should keep in mind I can have the same list of
elements in my rows and columns and you cannot have a different party. For example, I
cannot have architect, steel fabricator, d, e, h and so on. So, if you want to take teams it
should be completely teams, if you want to take parameters you should be completely
parameters in your rows and columns that you should not forget ok.
Then, talking about representations there are so many ways of representations. The
default one I think I have told you, it is primarily X mark between the cells to show the
relationship between the cells. I can have other forms as well. The same example
different examples sorry I have kept numbers as 0 1 this is called binary DSM and the
one generally represents relationships between the between the two entities ok.
So, this means K gives information to A, A gives information to B, A gives information

to F, I gives information to F and so on. So, I can still have a binary DSM on 0 and 1.
This is also acceptable. I can also have variables for showing my for explaining my
relationships. It can be weak, strong, moderate to show the strength of dependencies and
varying you can try eliminate the weak dependencies for making assumptions and then
you can go ahead in doing your schedules that is also possible, you can use variables,
you can also use numbers ok. These type of DSM is primarily called numerical DSM and
these numbers can mean anything to anyone.
This can also talk about strength of dependency. For example, E to B, it has a 0.5 as an
indicator on strength of dependency. So, wherever I have less values like for example, 1,
I can think of pushing this above the diagonals, so that even if I want to assume this I
may land up in very little rework compared to breaking a relationship something like this
or this with which you can still work it out and these are all the different ways of
representing relationships on to a DSM ok.
Now, let us move into operations. So, there are so many operations I have on the DSM.
So, one is a graphical representation and the formation of DSM I have shown you for the
same example how to form ok. Now, is as I told you this A till F shows a sequence of
execution. So, now, you will be very curious to know what is the nice sequence which
can have a minimal rework and so on? So, for which there is a process called
partitioning. Here if you see this block how many X marks are above the diagonal? Only
two X marks are above the diagonal and compared to four X marks as seen in this figure
ok. So, now, if you see the sequence of execution the sequence has changed from A, C,
E, F,….. A, C, E, F, B, D and so on. So, this sequence shows that I have altered my
sequence and so, that I have minimal X marks now above the diagonal ok.
So, now if you see here I have four X marks, here I have two X marks, still you can
really work it out on this is to see whether you can still minimize the X marks and this
entire block. Now, the shaded yellow portion is primarily called as a block. This block
means these set of activities like A, C, E, F have to work together in order to finish off
the whole work. That is what is the meaning of the block ok. Now, what I can do? You
need to really know where I can choose my X marks so that I can make minimal
assumptions and then I can work ahead in finishing my entire project. So, am I going to
choose two assumptions. Assumptions are here for example, this place and this place and
as a result my re-partitioning DSM is coming like this.
So, when I take these two assumptions. These two are the assumptions I have made. I am
starting my C, then A, C, E, F when you are plotting your values on duration on D, C, D,
A and D, E and for this sequence you will get a duration for what is the entire stretch on
the project ok. So, that is what is primarily the basic operations you do. Then the next is
primarily the iteration modeling wherein you plot the duration along the diagonal and
then you really get the duration for the entire block.
Now, generally what we say is DSM generally has four operations: one is a formation –
formation is not like a random case because a list of activities you identify itself is
challenging which you should not miss. You can still go ahead with the same WBS
structure and so on in identifying the list of activities there is still no harm, but if you fail
to miss even a single relationship between the activities what happens is this complete
partitioning and all will fail ok. You will get a different matrix at different a different
solution altogether and you would not be able to work on the solutions. So, formation
itself is very critical a task as such ok. So, let us see formation as a first step.
Next is partitioning, how to read sequence rows and columns simple excel we will be
able to do it and I will show it to you. Next is tearing. For different combinations of
assumptions, I will show you how are the different repartitioning happening ok. And, the
last step is primarily called so, every tearing is generally associated with the re-
partitioning and once a tearing operation is done then the last step primarily we call it as
an iteration modeling ok. Iteration modeling only you plot the duration along the
diagonal and then you work out as to what is the rework, how many times any activities
repeated, what is the rework duration assumptions made and so on. So, these are the four
steps we have in the DSM operations.
With this I am stopping for today’s class. We will see partitioning in the next class.
Thanks.
Indian Institution of Technology, Delhi
Lecture - 08
Dependency Structure Matrix – I
Title of this today’s lecture is on Dependency Structure Matrix. In the last class, I have
introduced and I gave you slight background on what is design management, what is
information and so on. So, these with the help of this a technique called DSM, you will
understand how to also go ahead in finding out scheduling aspects in DSM. So, let us
move on the contents.
In the last I have also shown you what are the basic operations in DSM like formation of
DSM, partitioning of DSM, then tearing process and iteration modeling.
So, in that first two operation basically the formation and partitioning is what we are
going to see in today’s lecture.
Just to brush up your mind. So, this is a formation of DSM, I have taken a same example
as shown in the last class. This is a partitioning where in for this particular example, I am
getting one block ok.
And then once you do the tearing process, in the sense identifying the tear marks or place
for assuming then you get the tearing block like this. And once you finish, then you have
to again do the repartitioning and you will see this block. So, primarily the three
operations are here; this is operation number one: formation, this is operation number
two which is a partitioning, the third one is tearing. Tearing always goes along with the
repartitioning, and then the next step is primarily the iteration modeling. So, now, unless
you stop with the three steps, you are only identifying the sequence outside the block,
with the partitioning you are identifying the sequence outside the block. And you are also
aware of where are the block formations ok.
In the tearing and repartitioning what you are going to do is, you are not touching outside
the blocks you are only going to operate within the block. So, within the block where are
the tear marks and what is my sequence within the block and what is my process of
execution that is what you are going to see in tearing. Now the question that comes to
your mind is, is there a better sequence than this; for example, C A E F and then B D. So,
primarily with the C A E F is there another alternate sequence? So, primarily for factorial
combination sequences are possible for the four activities within the block.
Now so, accordingly you have like 16 combinations are possible ok. So, one way is to go
ahead in evaluating all the 16 combinations and then seeing what is a best duration or
other way is primarily to do iteration modeling ok. So, how to test the performance of the
sequence, the only one method is add duration along the diagonals; and then find out the
duration for this project or for the block whatever it suits well. And then you have to
find, what is the, how much of rework has happened, how much of waiting time was
there, and what is happening, which activity is taking more time in the rework; so you
can analyze with the help of iteration modeling, that is a phase number 4 or operation
number 4 ok.
So, now as I told you, we are going to see with the formation of DSM. So, building or
creating a DSM, in other words we call it as a formation of DSM. I gave a very simple
steps, this is like it can go through for any way with which you are creating a DSM ok;
which also I will little bit I will explain later. So, first step is to define the scope of the
project ok. So, for example whatever project you want to take, how many our disciplines
are there, how many our components are there, you choose accordingly, and we define
what is the scope of the present project.
Then you decompose the entire project into various elements as possible. So, you can
stop at level 1, you can stop at level 2. So, the common method of decomposition we
have seen in the earlier classes is WBS, you can go through the same WBS methodology
in order to decompose a system. And you can stop at level 1 2 3 wherever you want, you
can still do the same and you can build the DSM; I will show you that also in today’s
class with various levels of the DSM formation.
Then identify the list of elements; elements is a common name which is given for either
you want to do component, teams, activities, parameters, whatever. So, when I am
decomposing a project, I may land up in so many different sub-levels which we have
shown in the last class. So, it can be components, teams, activities, deliverables,
parameters etc, so you can stop at any anyone place and accordingly you can create that
particular type of a DSM.
So, if you stop at activity DSM activity level, you can form an activity DSM; if you stop
at component level, you can form a component DSM. So, there are so many types of
DSMs also available. So, list the elements in the same order in the rows and columns.
So, accordingly what you have to do is, you have to identify the elements; the elements
is generic name given for components, teams, parameters and so on. So, if you want to
do out activities; then list out all the activities in the same order in rows and columns, so
that you are getting the skeleton DSM.
Skeleton DSM is a blank DSM with just the rows and columns identified and no
relationships are identified ok, that is primarily called skeleton structure. Then collect
data on the relationship, once you have listed out all the rows and columns; the next step
is finding out the relationship between those identified rows and columns ok. For that we
are suggesting a method with which we have worked out; and that you may follow or
there are several other methods with which you can also build or define the relationships
ok.
Then check for the validity, once you have done all the X marks it does not mean it is
fully complete, go through the experts and see whether you there is still any unidentified
relationship or whether there is a wrong identified relationship, something is missing and
so on; that you have to cross check.
And once you are very clear that you have done everything perfectly fine, then you can
really go with assurance that the matrix formation is fully complete. So, this is a very
basic steps with which you can create a DSM. Now let us go through the steps in detail
with an example.
The example, first example which I have taken here is IDCT, which is induced draft
cooling tower ok. This example it consist of four disciplines such as; civil, mechanical,
electrical and instrumentation. So, there are four disciplines, for the particular scope of
this IDCT project ok. So, that is the reason why I have chosen this.
Now, the cross-sectional details of this IDCT is given here or the induced draft cooling
tower is given here. I think many of you know what is a cooling tower? It is primarily a
part of a power plant project in order to cool the water which comes from the turbines
ok. So, now, what are the levels? So, primarily it has several levels one is; drive
assembly, drift eliminator, water distribution, fill level, sill level, cold water, basin,
foundations and so on. So, the water which comes, water is generally coming through the
hot water basins and it moves in through the drift eliminators and the water distribution
lids ok.
What happens here is the fans and the drift eliminator prevent the water particles from
going outside ok. So, this is primarily the hot water basin with which the water is falling
in and in the fill level there are the cooling medium ok, which cools the water. And once
the water is falling on at very large level, naturally the water gets contact in touch with
the air and the water gets cooled; and then it starts going into the cold-water basin where
the cold water is collected again for recirculation of the water process ok. This is
primarily the mechanism with which the IDCT works.
Now for the particular example in formation of a DSM, we are going to see only the
Drive Assembly level, which is the topmost level ok. So, little more zoom in view on the
drive assembly level, to show the sectional details. If you see here, so this is primarily
the, let me use my pen ok. So, this is primarily the hatch in the deck for all the operations
to happen and these are all the handrails are shown here, this is a ring beam, this is a fan
stack; fan stack is primarily in the cover to the fans ok, then we have the fan and the fan
blades.
Then this is a gear box mechanism which is here, and we have the fan stack door for
people to move out and inspections and so on. Then this is a motor, this is a pedestal for
the motor, and this completely you have a fan deck slab for people to move around. So,
here if you see there are all people coming in, you need electrical wires and cables and so
on, in order to move the fan, operate the fan motor and gearboxes.
Civil work is already there, these instruments are primarily the mechanical equipments.
And we also have so many instruments to check the noise on the fan, to check the
vibration of the gearboxes and so on. So, there are other instrumentation assemblies are
also there ok.
So, now what do I do, as I told you first is you have to get the list of elements ok. The
project is defined, I am going to work only on drive assembly level that is very clear;
then the end is decomposition level, decomposition I am going to use the WBS structure
only.
So, first is level that is how the project they are very comfortable in working. So, I am
also going to show in the same way. So, drive assembly level that is a level with which it
gets decomposed; then you have mechanical team, civil team, instrumentation team, and
electrical team, there are four teams in the particular project. And you know what are the
components here fan, motor, gearbox, driveshaft, accessories in the mechanical side.
For civil discipline, there are accessories, fan deck slab, fan stack and so on; In
Instrumentation, level switch, accessories, vibration monitoring systems; electrical again
accessories, elimination systems, and the earthing and lightning system ok, so these are
all the components. So, I have one-one activity identified for all the components, never it
is one-one; but in this case it happened to be one-one. So, I have fan design, motor
design, gearbox design, drive shaft design, details of equipment handling system, details
of removable walkway, handrails and trolley, details of ladders, doors and hatches.
Primarily the layout and all these has to be shown. So, these are all the list of activities I
have for the mechanical discipline ok, in the drive assembly alone.
In civil work details of insert plates, anchor, bolts for all the cable trays and so on; then
design and details of slabs, grid beams, pedestals at fan deck level; then design of fan
stack and ring beam. So, these are all the activities in the civil discipline.
Then for instrumentation data sheet of levels switch, instrument location and cable
routes, data sheet for vibration monitoring system, so these are all the activities we have
in the instrumentation. For electrical, cable trays and trench layout, illumination systems
or lighting system of the cooling tower, earthing details and lightning protection; so
accordingly we have almost like 16 activities in the entire drive assembly level.
Since it is very big an activity list, I have given an activity ID for the same starting from
D; because it is a drive assembly a level, so I have used D as a code. So, D stands for
drive assembly and A B C D are all the list of activities ok. So, I have used simple WBS
for the decomposition. The next is how do I, so with this I can easily get my skeleton
DSM and I can fill in my relationships.
But what happens is with the way you wrongly identify a relationship, or you put the
relationship in a different place or something or you miss out identifying a relationship;
you get a different result altogether. In order to avoid that, we are using a new
methodology. So, this methodology is nothing, but multiplying 2 matrices. So, that I get
a DSM ok.
So, these 2 matrices are variable DSMs or we call it as a DMM, DMM stands for
Domain Mapping Matrix ok.
So, the DMM is primarily the domain mapping matrix which stands for which actually
has a rectangular shape and it is not a square, wherein it maps relationships between two
entities. So, one can be as here, it is primarily a list of activities on the rows and along
the columns I am having a list of parameters ok. So, with this you can form all the
relationships and what happens here, any activity will have a set of outputs at the end of
the activity execution and activities also need inputs from other activities ok.
So, every activity needs a set of or list of inputs. So, when you multiply these two you
will actually get an activity DSM, that is what is the main purpose behind. And logical
way of doing this also I will show you right now, manually also you can do, you can
make this as binary matrix and multiply the 2 matrices you will obviously, get another
matrix ok. So, for example, let me read through. So, fan design, so this is for output
parameters and this is for input parameters. Fan design can release fan power
calculations which is required by motor design and gearbox design; and fan design can
release hub and seal-disc diameter which is required for design of ladders, doors and
hatches. So, fan design also releases as fan diameter and tip clearance which is required
for numeration and RC details of fan stack and ring beams. So, like this you have to go
through this.
So, what happened is? So, I am having a DSM like this. So, these are all my activities, I
wanted to form an activity DSM. So, DA, DB, DC, and so on; as I told you for beginners
it is better that you write what is the X really representing for, till you go for validation
and check for the whole flow. Once the whole flow is really defined, then you can cut off
all the values and you can replace it to X marks, which is very easy to do.
For example, so DA is releasing fan power calculations, which is required for motor
design and gearbox design. So, what I have written here is DA to DB, what is the X
actually there, it is a fan power calculations, DA to DC what is a X really representing it
is a fan power calculations. Now DA also releases hub and sealdisc diameter which is
required for DG, which is details of ladders, doors and hatchers. So, same way DA to
DG hub and sealdisc diameter is the parameter which is traversing between from DA to
DG. Now same way the fan design also releases fan diameter, tip clearance which is
required for activity DJ which is a numeration and RC details of fan stack and ring
beams.
So, what happens is DA to DG, primarily the fan diameter and tip clearance is what is a
input which is travelling from the one activity to other activity. So, same way, we can
travel through the entire activity DSM; motor design releases motor rating which is
required for gearbox design which is shown here. Now this motor design also releases
selection and sizing of motor which is required for activity DI, which is the layout
numeration and details, so here you have written here, selection and sizing of motor.
The gearbox design releases an output called selection and sizing of gearbox. The
selection and sizing of gearbox is actually required for motor design, it is required for
details of ladders, doors and hatches; it is also required for layout numeration. So, it is
required for three activities. So, selection and sizing of gearbox is required for DB DG
and DI.
Now the details of ladders, doors and hatches releases information on fan stack door
details, which is actually required for activity DJ which is the numeration and RC details
of fan stack and ring beams ok. So, which is written here and accordingly the entire
matrix is filled. So, this is how the activity DSM is filled.
So, now I have got the same matrix, I have shown it here. Now once you take it to
validation, it is easy for somebody to go through ok, fan power calculations this is
required for here. So, it is easy for somebody to validate this matrix and once a matrix is
validated then you replace it to X marks for your further calculations ok. So, now,
formation of activity DSM is really done ok.
So, so far we started up with defining the scope of the project, then decomposing the
project that I have shown you the levels ok, teams, components, then activity list; and
then use variable DSM, so the other name I can say is DMM which is domain mapping
matrix. And I am having a rectangular matrix 2, rectangular matrix come; then primarily
you have to multiply if you want to mechanically do or you can go through the logic also
I have explained you that, you do that. And then you are getting this matrix convert into
X marks, so that you will get the activity DSM ok. So, which is what I have shown.
Now in this, define the scope, define the decomposer system. So in this decomposer
system I can stop at various levels. In the previous example I have stopped at activity
levels, now in the next example I will show you how to form a team DSM or a
component DSM using the same methodology.
Now whatever you have stopped you wanted to form the type of DSM, you can
decompose elements in the same way, list the elements in the same order and you can
still collect data on relationship and you can still go ahead with the matrix formation.
Now, I am going to take the next example; the next example is on interceptor sewer ok.
This is a very simple example; I have shown this example in different form in the earlier
classes.
Primarily I have four components ok, I am going to work at component level. So, I am
only showing the components, I have intercepting chamber, manholes and sewage
pumping stations, lifting stations are also there, but I am not showing right now in this
particular work, and this red line is primarily the intercepting sewer ok.
Now, for the particular Manhole component alone I am going to show the team DSM
formation ok. So, now, these are the 3 teams T1 T2 and T3 and these are primarily the
drawings D1 and D2 and D3 are the 3 drawings ok. If you see here, the identification is
given in the legend here, topological survey team is T1, geotechnical investigation team
is T2, structural design team is a T3 and I have details on deliverables.
So, D1 is location details of manholes, D2 is GA of manholes, GA stands for General

Arrangement and D3 is design of manholes. So, this is a first DMM which I have formed
with output deliverables. So, T3 is actually giving two deliverables which is, D1 and D2
ok. Now these deliverables are required, there are several deliverables required for the
teams ok. Let us see on the input deliverables, so D1 is required for T1 and T2, D3 is
acquired for T2 and T3 ok. So, now, if you see here, I am going to form a team DSM;
now I have not used all my X marks here, I will tell you that also later. So, T3 is
releasing an input called releasing an output called D1 which is required by T1 and T2;
T3 is actually having a dependency relationship on information with T1 and T2 ok. So,
this DSM is done ok.
So, I have used only the few relationships here, other relationships I have not used. So,
which is what I said, when you actually avoid this X mark and then write properly what
is it; like I showed you in the last example, it is easy for you to map the complete or go
through the complete flow of the process, design process and map all relationships ok.
And the other problem that happened here is, I have used a small example so many of the
relationship is got missed out, when you are really building up a complete a design
example all relationships can be easily be identified and shown ok. But the purpose of
this example is to only show the, how to form a component sorry team DSM for the
manhole component ok; same manhole component how to form a deliverable DSM ok.
So, I am going to use DMM again, but there is a list of deliverables and output
parameters.
So, now, these are the deliverables same the deliverables I have location details of
manholes, general arrangement of manholes, design of manholes; and I have identified 8
parameters for this all these 3 deliverables. Pipe diameters P1, P2 is diameter of
chamber, P3 is diameter of axis, P4 is height of the chamber, P5 is grade of concrete, P6
is type of reinforcement, P7 is coefficient of earth pressure, and P8 is dead load. So, I
have all these parameters, same logic still works good. So, I have list of deliverables, I
have list of parameters here. So, these are all my X marks identified for the parameters,
same way for the input parameters.
So, now, what happens here is, I am going to use for my D2 is actually releasing P1
which is required for D1 which is identified here. So, D2 is releasing P1 which is
required for D1 which is identified here ok. So, now, D3 is releasing P8 which is
required for D2 which is identified here ok. Other parameters did not come within the
scope, but the idea is to convey to you. So, the logic for formation, if you want to
identify a complete DSM is, go through decomposition, WBS will still hold good for you
and then formation of DSM in, so sorry. So with the help of WBS you can identify all
these levels ok. It can be level 1, level 2 whatever it is, drive assembly level or you can
go through team level or component level, activity level. I have shown you several cases
right now, so you can stop at any level. WBS will help you to identify only this, once this
is identified for identifying X marks you can multiply the variable matrices and you will
still get the relationships identified. Then go through with the experts, see whether
everything has been identified and then you have to validate your matrix formation ok;
because if you identify a wrong matrix then obviously, your results are all going to fail
very bad ok. So, that you have to take care.
So, that is all on about matrix formation. So, now, the next step is Partitioning ok. So, as
I told you first step or first operation in your matrix operation is formation of activity
DSM.
Once the matrix has been formed, the next step is primarily the partitioning processes.
So, what is partitioning, I think I gave a quick info earlier also while I introduced in the
lecture. So, partitioning helps you to identify what is the relationship of the entire project
and wherein there are no interdependencies or cycles or let me put it another way. So,
when the project has independent and dependent activities those are all identified in the
partitioning process and wherever they are cycles, the interdependence, they are all
group together as a block formation.
So, in a complete project, I will know what is a sequence of my independence and

dependence which are all collected together and then I know where is my block
formation happening ok. And the methodology also I will explain, which is a very simple
method on matrix multiplication; I will show you that and along with an example, so that
you can follow me along with example and various steps how the matrix gets changed
ok. That is all is about the partitioning process; partitioning never talks about relationship
within the block and also it never talks about sequence within the block ok. This A C E
F, so I have shown here, partitioning will not talk about this sequence; so you have to go
through the next step in order to identifying what is the sequence there ok.
So, let us now you see the definition of partitioning, it is a process of reordering the
DSM rows and columns, it can be component, it can be teams, it can be activities, it can
be deliverables; whatever you have identified you are only going to reorder in such a
way that the new DSM arrangement does not contain any feedback marks or primarily
you have to minimize it into lower triangle ok. So, thus transforming the DSM into your
lower triangular form ok; so primarily I will be getting a lower triangular matrix ok, if I
have only an independent and dependent relationships. Now what happens is, if it is
difficult to make a lower triangular matrix because of the interdependencies and cycles,
attempts are primarily made in keeping those elements identified as a block formation
ok. So, still there are methods which can make you to keep these X marks close to the
diagonal, but partitioning never attempts on that ok.
In doing so only few elements will be involved in the cycle, so try to minimize this
formation in self identification of block itself. Now what is the meaning of this block
formation, that also you should know. So, this implies these four; suppose if these are all
activities, these four activities have to be work together till the completion, so that
because each of them reads input from each one of the other ok. For example, ACEF is
primarily a block ok, what is the meaning of the block? The teams have to work closely
together like this in order to finish the completion of the particular scope of the work ok;
and then activity B and D. So, that is a meaning of the, logical meaning of the
partitioning block formation ok.
Now, what is the partitioning flow process? Now there are several methods for
partitioning as given in the websites ok. If you go through the DSM website as such, you
will be there are so many methods like path searching, matrix multiplication, or we have
triangularization matrix like this n number of methods available ok.
So, for identifying the sequential the parallel and sequential activities, topological sorting
is the best way to do. And once you identify the topological sorting and identifying that
there is a big block, but I do not know what are all there within the block, then you have
to choose one of the methods like path searching or matrix multiplication and so on ok.
But it is not a straight forward method, you have to work in cycles; primarily there is an
iteration inside, which with an example. I have taken an example in such a way that, it
can cover all the problems which you have in the examples ok. So, within this cyclic
process also I will explain properly.
Now partitioning flow process, there are three broad items here; one is topological
sorting, primarily to identify all my sequential and parallel activities. Then I do matrix
multiplication, matrix multiplication or take off this concept and put path searching can
help you to identify or trace the cycles only; once you have trace the cycle, then how to
group them into blocks again repeat the process on topological sorting and matrix
multiplication until you identify all the cycles and all the activities ok. And at last you
get the partitioning DSM; this is the overall flow on the partitioning process ok.
Now, let us go through with the hypothetical example. So, for explaining this
partitioning process, I have taken a hypothetical example ok.
This is a generic A B C D example, I have taken A till K as a simple example; because I

have to show all these n number of steps, I have taken a little bit bigger example only.
What is the first step as shown earlier, check for X marks above the diagonal? Do I have
X marks above the diagonal, yes I have lot of X marks above the diagonal, which implies
this is not the right sequence for you to execute the whole block. Suppose if there are no
X marks above the diagonal. So, what it means is, the matrix is already partitioned ok,
already you know what is the best sequence to operate the entire flow ok; that will be
very sure if you do not have any X marks about the diagonal. Then directly you can go to
the last step in saying the matrix has been already partitioned ok.
Now, since I have X marks, so same example I have shown you in a smaller form in the
top, so that you will follow the slides.
Now as I told you, once we have X marks above the diagonal; the first step is to do
topological sorting ok. Topological sorting is the graph theory terminology or concept.
So, what does it say, I have written very simpler steps for you. So, primarily you have to
identify all empty rows and columns or rows and columns in order identify the start and
the last activities ok. So, when I am having a starting activity; obviously, I would not
have any X mark along the rows. So, that will be shown as an empty row and when I am
at the last activity, then I am not going to give any inputs to other activities. So, I may
have empty columns and, but I may have lot of X marks along the rows ok. That is the
meaning of topological in order to identify the starting parallel activities and the last
activities. Primarily the first activity, last activity, I am going to trace out now with this
method ok.
So, first step is move all the empty rows and the corresponding columns to the top and
left respectively; and omit those activities for further consideration ok. Now, let us go
through this example, where are the empty rows right now; activity D has an empty row.
So, there is a complete empty row here ok, there is nothing given in the activity D. So,
what I am going to do is? So, I have identified that activity D has an empty row, you can
use excel sheets for doing or there are software also available which can help you to do
the partitioning process ok. So, I have moved my activity D to the primarily to the top
and left, because I have to pull out both my rows and columns accordingly outside. So,
just cut off that row and the particular column and move to the top and push all the other
activities. Then you will obviously get this matrix. If you want to do manually; otherwise
you can create you can run a small VBA program and you can still go through the
exercise ok.
So, now, what we have to do is, omit those activities for further consideration. So, in the
first time I have written, move them to the top and left and then do not include them in
the further for the calculations ok. I have identified D as my first activity, and I am going
to ignore this in my further analysis ok. Now what you have to do? You can repeat the
process of identifying empty rows again or you can go to the other step as identifying
empty columns and you can swap this up and down ok. In this particular example, what I
am going to do is. I am going to identify all the empty rows after omitting the activities;
and at the end I am going to identify my empty columns and this I am going to repeat it
again and again ok.
So, now let us see the next step. So, what I am going to do is? I am going to identify all
other empty rows and rows also. Now if you see, after omitting this activity where are
my empty rows, this activity E has completely no X marks along the rows. So, activity E
is primarily my second activity in the list ok. So, D is giving inputs to activity E, but it is
not the first activity, but it is a second activity ok. So, push the activity E next to activity
D and then omit this from further consideration ok.
Now do I have any other empty rows? So, in this example I do not have any other empty
rows. So, primarily I have finished my first step on: move empty rows completely; which
primarily is the starting activity or next sequential activities ok. I have identified all the
first few start activities ok; after then the block starts, and you are not aware of where the
block is and so on ok. Now I am going to move this. So, I am going to finish this step in
cycles and then I am going to do this. But as I again I am telling you if you want you can
swap these two in order; you do this first, do this second, then you can do this. Primarily
what I have done is, I have finished all, this step is over now, I am going to go to the next
step now.
This is my last step I have shown this picture in the top corner. Now move all empty
columns and corresponding row to the right and bottom respectively and omit those
activities from further consideration ok. So, what are the activities that are empty; I have
activity C which is empty here, I have activity J also which is completely empty ok. This
looks like C and J are parallel activities; but they are in the last in the sequence ok. So, I
can move C and J completely to the last. Now what is the order of C and J, that is there is
no restriction here; because both are empty, both are parallel activities, there is no worry
if you want to place C first and J first or you want to place J first and then C first, there is
no problem at all ok.
So, now, I have moved C and J to the last. So, now, what happens is, I have identified D
as my first activity, second is E activity, I have also identified two parallel activities in
the last as C and J. Now again I have to repeat this process to check any other empty
columns are there. So, any other empty columns are there, it looks like there are no other
empty columns ok. So, this process I have to repeat in cycles, which is what I said ok.
This has to go in cycles, until completion. Now within this then have got a bigger block
and all empty rows columns are identified ok. Only the start last activities are identified.
Now you have to find out what is there within this particular stretch ok.
Now, what is the next step? So, topological sorting process is over ok. Next step is,
matrix multiplication in identifying cycles. So, now, what you have to do? I have got this
matrix, now I am going to give a new name for this matrix. So, this is primarily the
active matrix, which implies the matrix for consideration is only this portion for you,
because you know D is first activity, E is second activity, C and J are the last activities
that is known to you very well.
Now your main botheration is to know what is a sequence within this, ok. Now what you
have to do is, convert the active matrix into a binary matrix. Now this is my active
matrix, now how do I convert into binary; while explaining the types of DSM I have told
you this. So, replace all these X marks as 1, if you are working on excel again it is very
easy, select for all these this particular cell ok, search for X and replace with 1 you will
get all ones; empty cells all if you put 0, you will easily get a 0. So, I have just converted
all these into a binary matrix. Now I have my binary matrix. So, this step is really over
ok. So, take out the active matrix and I have converted into binary and I have the binary
matrix right now, ok.
Now, what do I do? Raise the powers of the matrix.
So, this is my example. Raise the powers of the matrix, until there occurs a non-zero
entry along the diagonal anywhere before power n or else stop at power n. Power n, n
primarily I am using an alphabet which talks about the number of activities here ok. For
example, in this case I have 7 activities, so I have to stop till power n, whatever happens;
or well ahead of power n, when I am able to get some non-zero along the diagonal ok.
So, I have raised, I have a binary matrix which I have shown you in the previous case.
So, I have done primarily one multiplication only, which we call it as power 2. Power 2
is the binary square ok. So, I have done multiplying only once.
And what happens is, I identified a non-zero entry here ok. So, this implies G is here, H
is here and I here. If you are not very comfortable in identifying or tracing the loop, you
have to do the tracing by go throw the relationship and then identifying where is the real
loops ok, because I have got values as 2 also. This 2 implies this H is on two cycles and
G implies it is on one cycle and I implies it is available on one cycle. So, how does this
relationship look, you can go through later after cross checking with the example. So,
this is what is the tracing on the whole loop. So, this is primarily we call it as
overlapping cycles. So, H is primarily is in the midway between G and I, in the length of
cycle of length 2 ok; that is what we have identified in the matrix square or adjacency
matrix raise to powers 2 ok.
So, now, what happens? Once you have identified the activities corresponding to the
non-zero entry, I identified it G H I are in those activities. Check for the individual loops
that also I have done, sometimes I may get values here there, this maybe in some other
block; this need not always be together in one block only it can be two three blocks of
cycle of length 2 can also happen. So, you have to check for individual loops; the one
way to do, again I am repeating go through with the original example and trace for the
cycle, so that you will get the cycle ok. Then what you should do? Collapse or merge
those activities on the loop into one. So, what I have done G H I since it is all in one
group only, I have merged the G H I into one block. So, I have made it as one and I have
merged the relationships also into one ok. So, I have now the new matrix which is like
this ok. Now what you have to do, again you may have.
So, let me also tell you what is the problem? If I have not stopped at this step, ok. You
should also now, you may have a question why should I go through this merging and this
process again and again; why cannot I go till power n directly and then identify all the
bigger cycles or loops. You would not be able to identify; for example, I will tell you this
is raised to power 2 ok, the same example I am raising it to power 3, I am getting A B F
as like this and then H I K as something.
So, the minute you go on without merging the first formed loop ok. So, the first time you
are getting a non-zero entry you have to stop there; and then you have to identify the
cycle and then proceed after merging only. If you do not do, there is no use in the matrix
because these loops will give new values and then you would not be able to identify
properly the loops, ok. So, that is what I want it to show you on what happens when you
raise it on power 3, you can cross check with any example later ok.
So, now let us move on. So, what you have to do after this. So, once you identified the
loops, you got a smaller block within the bigger group; and what you have to do is, you
have to again do the topological sorting ok. So, which is what I have written here, so
merge them into single activity; then repeat the topological sorting, then check for the X
marks above the diagonal; topological sorting generally talks about that only.
Identify the highlight the condensed loops as blocks, if there is nothing else available ok.
Now this also goes on in cycle. Let me show you with example then you will understand
better. The same example as shown in the previous slide I have taken it up here; now
what you have to do, there are X marks above the diagonal. And what about topological
sorting and I do topological sorting, I do not have empty rows or columns ok; because
you merged a block completely into one, you may also get empty rows and columns now
ok. There maybe small blocks within this block also, you have to identify everything ok,
so that, you will identify when you are doing the topological sorting again and again.
In this case, nothing happened because there is no empty rows or columns are seen in the
example; then what you have to do, you have to identify the remaining cycles or loops
ok. So, I am converting into binary, same process ok; then when I raised it to power 2
identified that, there is a block of cycle length 2. So, F is dependent on G H I, and G H I
is dependent on F which is what I know ok. This if you are very an experienced and not a
beginner, you will also identify this with the help of this X marks also. If you see here
with X mark here it shows that these two activities are interdependent ok; you need not
even do the multiplication, without that also I will understand; but if you are new to the
whole process better you do all the steps.
So, this I have identified, because I have got 1, 1 of this implies F G H I or in cycle. Now
what am I going to do, I am going merge the F G H I as one. You will see here the
merging, primarily I wanted to show both the merging together. So, if you see here since
I merged the F G H I in to one, you will see the other relationships are not missed, but
this relationship within relationships within the F G H I are all got done in the sense they
all are not visible ok; that is the meaning of showing the merging.
So, K is, so K gives information to A that is seen here, A gives information to B that is
also seen, A gives information to F or F G H I that is also seen here ok. Now F gives
information to G H I and G is giving information to F G H I you need not show, because
there is no relationships which is done.
This you need not even worry ok; when you are merging it, just merge both the X marks
together then obviously, you will get all the X marks identify ok. Now K is giving
information to K to F and to G H I which I have shown it up here ok, now B is giving
information to K that is also coming here, which is also seen here ok. So, all
relationships you have to identify, the relationships within the particular activities in the
merging you need not show. That is the meaning of what I am saying as merging into
one; because you know that there is a block, so you wanted to keep it as one activity and
then start repeating the process ok.
So, now I am going to repeat the process. So, I have four activities right now A B F G H
I block and then activity K. Now when I repeat the topological sorting what happens, see
here the F G H I block is primarily in the last and A B K has come to the front ok. Now
there is a cycle within this A B K which you have to identify; because in the further
topological sorting what happens, there is no empty row or column so obviously, you
know that there is a block in the A B K ok. Again, if you are really experienced with the
DSM you will understand.
So, there is a cycle of length 3 here, because I have this pattern here. K gives information
to B, A gives information to B, B gives information to K. So, primarily A to B to K
which when you are raising the matrix into power 3, we will get one along the diagonal
you can cross check that later. So, you will identify it as power 3 and when you are
raised to power 3 and there is a cycle of length 3. So, now what you have to do, merge
this A B K as one activity, F G H I as one activity; when you are merging you are getting
a relationship here ok, during the merging.
So, this automatically is topological sorted ok. Otherwise you may have to repeat the
process again. The automatically I am having an X mark below the diagonals, so this
implies the matrix is sorted already ok.
Now, where did we start, we started upon this: the topological sorting process identified
activity D as my first activity, E is my second activity, C and J are my parallel activities
in the end. So, start of the project, end of the project I have identified and there are lot of
activities within the middle ok. That you wanted to really explore that, for that we have
done Matrix multiplication; we identified that A B K is one block and the F G H I block
is dependent on A B K. If you see here are all the 7 activities in the block has been
identified, yes I have really done.
So, what is a partitioning DSM I am having. So, D E is my first, A B K block is here, F

G H I is here, C J is here, so this is my sequence, these are the two blocks I have got ok.
So, that is primarily the steps in partitioning. It may look like you have so many steps;
but I have explained you the logical meaning behind those steps also, you need not go
through all the steps if you are very comfortable in seeing the relationships and working
on with DSM. If you are very comfortable in plotting out the relationships on your own;
then you can easily identify all the blocks, and then plot the X marks ok.
So, that is all on above the partitioning example. So, so far we have seen formation of
DSM and the partitioning of DSM in the basic DSM operations ok. So, I have taken two
cases; one is formation, one is partitioning. In the next class we will see the next two
operations in DSM; it is primarily the tearing process, the repartitioning of the tearing
processes, and iteration modelling ok.
Thanks bye.
Lecture - 09
Dependency Structure Matrix-II
Title is Dependency Structure Matrix again and we are actually continuing with the last
class. So, last class we have seen the two basic operations; predominantly the formation
of a DSM and the partitioning process. In this class you will learn about tearing process
and the iteration modeling which is a third and the fourth operation ok.
So, what is tearing process, so let us see. So, it is a process of choosing a set of feedback
marks. So, primarily feedback is the name given to X marks which are above the
diagonal and feed forward marks are the name given to X marks which are below the
diagonal. So, in a way if I take this example so all these X’s are primarily the feed
forward and these two are primarily called as a feedback marks. So, you have to choose a
set of few marks, if you start choosing all X marks then you may have to repeat
everything every now and then.
So, you have make the number of tear marks as less as possible ok, as minimal as
possible. So, it is a process of choosing the set of feedback marks that if removed from
the matrix and then the matrix is re-partitioned which will make it as a lower triangular.
So, the marks that we remove or generally are called tear marks. So, in this partitioning
example I have this example in graphical form also and these are the feedback marks
which I am choosing to remove and how do we do; one way is to analyze the network
like how I will show in the next slide is to analyze and network and choose a set of
marks. Even then it is better that we have NDSM.
NDSM is nothing, but Numerical Dependency Structure Matrix. While I was explaining
you with the various representation on DSM. I little gave an introduction on numerical
dependency structural matrix. So, this number can be any variable which can range from
0 to 1. The number replacing the X marks can be 0 to 1 and it can mean to anything
whatever you want to keep. It can be strength of dependency, it can be sensitivity on
information, it can be information what does it called dependencies, or it can be even
reverb probability, it can be anything that you wanted to keep between these two
activities ok. So, primarily use them in order to choose the marks for tearing process
based on the type of project.
Now, let us see a little more on the analyzing the network. The age-old technique which
people used for analyzing the network, primarily the cycles or loops is called as a Shunt
diagram. So, let us see that little more and then we will go on into NDSM for choosing
the set of feedback marks ok. So, I am having a partition DSM. The same example which
I brought it up in the earlier classes, I am using the same example so that you can easily
follow till the end and the same graphical representation I have arranged it here for this
sequence like A C E F and B D.
So, this is my block representation and then I have to do my B and my activity D. So,
that is what is the real case ok. If I analyze this properly what I can see, I can see two
cycles of cycle length 2 ok. There are two number of cycles A C to C A and C F to F C.
So, predominantly there are two different cycles and cycle length is primarily 2, which
implies two activities are involved in the cycle of the loop. That is a meaning of the
same. Now if I little more analyze the network carefully, I have two cycles of length 3
also.
So, CA to F to C so, this is the link. C to A to F to C this is one cycle and the other one is
C to E to F to C this is another cycle. So, I have two different cycles again of cycle
length 3. So, within a block of four activities how many cycles I have? I have four
cycles; two cycles of length 2 and two cycles of cycle length 3. So, I have 4 cycles in
total if you see carefully. So, what you should do a common thing is choose any set of
feedback mark by attacking that feedback I am actually you know removing all of my
cycles or many of my cycles.
So, if I choose this X mark which is from F to C ok, if I choose is X mark which is from
F to C which is this so, what happens which is what I have shown here. So, what happens
is I have removed my C to F cycle. So, I have broken down this cycle ok. So, I have also
broken down this cycle which is a C E F cycle that is also has broken up and I have also
broken down the A C F cycle also ok. Three cycles I have broken down so, all three
cycles are gone.
If I have to choose one more tear mark in any of the two, then with to 2 tear marks I am
actually I have broken down all my cycles. So, I will have a sequential or a parallel
activities only right now to start with. Then repeat wherever you made assumptions to
check and then based on that you will get the duration for the entire processes ok. Now if
I have chosen another mark for example, I have chosen exactly a difference set of
feedback mark, so this cycle is broken.
Now what about this loop is still remaining so, I have to choose one cycle for here and
here also I have to choose one more, because there are two cycles here. One cycle of
length 2 and one more cycle of length 3 is here. Again, if I have to choose this, I have to
choose one more so, I have I have chosen this as a common cycle ok. So, minimal tools 2
tear marks you have to do or it can be 3 or even more ok that is what you have to do.
Now which is the best option is it this or this how will you tell? With number of
assumptions made; obviously, this stands out to be the best, but you do not know that
without checking out what is the duration of all these activities, what is the relationship
between those activities ok. Is it a really sensitive project or is it like very uncertain a
project every time changes are happening and so on.
Then there is no harm in redoing something and then as information starts flowing in you
can still repeat and go ahead. Then this situation may be better if your durations are all
smaller and so on. So, you do not know that without analyzing your NDSM. So, the need
for NDSM is really essentially in order to choose a set of feedback marks for breaking up
the cycle and a place to start for your partitioning activities ok.
Now, what happens as I told you earlier, every time you choose a set of feedback mark
so, I have chosen the common one as from the earlier case I have chosen the common so,
which actually breaks down three of my cycles. So, among these two’s I have chosen
two options. So, one option also I will show you, the other option also I will show you
and you have to again do the re-partitioning process. In this case the block remains as
same the sequence for the block is still ACEF and then BD.
Suppose if I change the set of feedback mark and I have chosen this, but this is still the
same, because three cycles are already gone. So, I am going to choose this as my set of
feedback marks, then in that case what happens is, meanwhile re-partitioning the DSM,
the sequence within the block is changing now to CAEF ok. So, this is what you have to
do with the help of re-partitioning.
Now I am just wrapping up from the start till the end in formation of the DSM. In
formation of DSM again you had two steps; WBS in order to get the list of entities for
example, list of activities and you used multiplication of two DMM’s in order to get an
activity DSM just to get the X marks. Still you have to verify with the experts and go
ahead and getting up refined your processes and then you are going into partitioning. In
partitioning it identifies all your sequential an independent activities which implies a
dependent and independent activities and all your blocks are all identified ok.
Now within the block, the sequence of the block you will not know in a partitioning step.
So, you have to do tearing. Tearing is nothing but identify in a sequence within the
blocks by choosing the set of feedback marks. So, I have chosen few places for tearing,
primarily for making assumptions. So, I have made an assumption here, I have made an
assumption here and then this is my flow. Once this is fully done then I have to repeat
the steps again and again in order to finish up for the assumptions made in order to
compensate for the assumptions made ok. So, then you are actually getting out the
sequence within the blog.
So, as a result what has happened you are a from the partitioning and tearing process,
you actually got the sequence of the entire project execution ok. You got the data
formation, partitioning, you got the overall stretch or a sequence of the whole project
along with block identification. Tearing you are actually finding out the sequence within
the block, you have done tearing process. Now the next question that comes is: is this a
better sequence or is that any other sequence which is really really good ok.
To answer that question then you have to really do iteration modeling, which is implies
identifying the duration of this particular blog. So, go through this, plot your durations
along the diagonals ok. Add values along this primarily the rework probability, what is
the percentage of rework or what is the probability for any change in activity A whether
it affects activity C or something that you may have to redo it. And, then see what is the
duration for this entire project or what is the duration if I go through this particular
sequence of a project execution.
You have to really check that cross check and then you choose the tear marks and then
go ahead ok. If you are not happy in doing iteration you can still go ahead with values
here and then doing it. Suppose if you want to really evaluate whether it takes 16 days,
this takes 20 days or this and other sequence take 18 days. So, what is it how much of
rework is happening in the activities, individual activities and so on then you should
really do the iteration modeling.
So, this dependency structure matrix. It is a little tricky concept and complex concept.
So, I thought I will introduce you with one more set of examples, so that you can
understand this better.
So, this is a hypothetical example. So, this first table is primarily a DMM on output
parameters ok. So, A, B, C, D, E maybe this can be teams or components. This can be
parameters or deliverable. So, whatever you want you can keep ok. And this table is
primarily on input items which implies. Let us assume if these are my teams or
deliverables. And these are my parameters. Then this is and DMM on teams and
parameters on output parameters. This is on teams and my input parameters. So, now,
how do you read this? So, primarily team A gives one parameter P1 team B is
responsible for releasing one parameter P2. Team C for releasing two parameters P3 and
P4 like that you may have to read.
Same way for the input I need P2 for completing my 5 team activities, I need P1 and P3
for completing my team activity B and P5 also. I need P2 and P6 for completing my
team C. I need P1 and P7 for completing my team D and so on. So, this is how you read
on the input parameters ok. Now, let us see how this is merged as a team DSM.
So, first you give A, B, C, D, E; A, B, C, D, E. So, this is primarily the list of teams I
have ok. So, A releases P1 which is required by B and D; so, A releases P1 which is
required by B which is required by D. Now, B releases P 2 which is required by A and C
ok. So, here if you see this is how it is. So, B releases P2 which is required by A and C
like this you may have to do P3, P4, P6 and P7 ok.
Once you cross check whether there are any omissions in the list of teams or parameters
and so on. Then once you have very fully comfortable with the flow. The flow you can
easily look at like this and then read giving and gives information to. Once this is
completely settled and satisfied, then you can go to merging it as a conventional activity
DSM by replacing all the variables that you have given here into X marks.
This is how the team DSM is formed. Suppose if you are very comfortable in the list of
activities relationships and you think you have done several projects earlier and you are
very experienced and comfortable. Then you can go in doing this directly in the first go
itself ok. Again I am repeating you can ignore. You can ignore the 1, 2 and 3.
You can ignore these 1, 2 and 3 steps and you can directly doing this formation of DSM
on one go itself. Suppose if you are very comfortable in understanding on the whole
project ok. This is primarily on formation of DSM. So, the basic operations in DSM,
number 1 is formation, number 2 is partitioning.
So, we will see the partitioning in the second step. And this is a generic example I have
taken. So, I have done the formation using any of the method I am comfortable as
discussed in the earlier slide. So, this is my original DSM. Now, I have to know what a
sequence of executing this particular example is. That is the main concept in partitioning.
Again I am just making you recollect: partitioning is resequencing the rows and columns
in such a way that, you can identify the independent and dependent activities and also the
block formations that is the main objective of partitioning DSM ok. Now, if you see here
to do this primarily you have to identify the independent and the dependent steps in the
first go. So, how do you do? You have to literally see whether there are any empty rows
and columns. First I am going to look for empty rows ok, where are the empty rows.
For example, activity D has a completely empty row here ok. So, what you have to do is,
empty rows implies they are the starting activities. So, move the activity B on to the first
column and to the first row and ignore the activity B from further considerations ok.
Keep it as highlighted and ignore this completely ok. So, I have ignored that and I have
pushed B completely.
Now check whether any more activity has an empty row apart from activity B ok. If you
look here there is no other row which has an empty cell throughout, empty complete row
ok. So now, the independent activity on the starting phase we have identify it is only one
activity which is activity B. Now what you have to do is look for empty columns again.
If you see for the empty columns, I have activity F. And I have activity H. Both are
empty cells there are no X marks in both the activities.
So, what I have to do is, I am going to push H and F to the last. Any order you can keep
because they both are parallel activities. So, I am going to keep H and then F and I have
prevented them from any more evaluations ok. So, now, is there any other empty column
again there are no empty columns no empty rows. So, primarily what I have identified is
B is the start activity H and F are the finish activities that much I have identified.
And I know there is a loop along with the interior part ok. Now, how to identify the
loops here. There are so, many methods for doing it and I will show you one method.
Primarily the block is very simple. So, I am going to show manually, otherwise you can
also use power of adjacency matrix on excel sheets and you can still get the result ok.
Now I am going to show manually only.
So, primarily A to C and then A to E then C to D ok, then D to C ok, then E to A and
then G to D ok, so G to D ok. So, I have identified all the activities E to G, E to G and
then G to D ok. Again I am checking. So, A to C is done, A to E is also done C to D is
done then D to C is also done E to A is done ok, then E to G is also done and G to D is
also done G to D is also done.
So, I have identify everything. Now if you see here I can see that there are two loops
here. There is a loop of cycle length 2. C to D and A to E ok. So, what I am going to do
is A and then so, primarily A to E so ok. So, this is sorry ok. So, I have two loops. So,
AE is one loop and then CD is one loop and then I have activity G and A E G afterwards
I have activity G.
So, primarily I have a block of cycle length 2 and then I have one more activity also ok.
So, this arrow mark is actually from G to D only. So, this is primarily the merging of the
blocks which we do ok. So, AE is one block, then I have C D is another block and G is
an activity giving inputs to CD and after AE I have my G.
So, how do I represent and the normal sequence? So, AE to G, AE to CD and then after
G, I have CD. So, easily you will find the sequence when you are evaluating in a
graphical format on the inner side ok. So, accordingly what happens is. So, I have my
block here. So, activity B is a first activity H and F for the last two activities. A E is one
particular block as I have highlighted.
And CD is one particular block as I have highlighted. And G is primarily the activity in
the middle which is giving inputs to CD, which is receiving inputs from AE and it is
giving inputs to CD ok. So, with this I have identified how to execute after B, AE has to
be done together. Then activity G, then CD has to be done together then H and F can be
done in parallel; because I do not have any X marks in the between H and F.
So this is all about partitioning method. So, what you have to do is, first identify the top
starting activities all empty columns will identify that. Once you merge one activity see
whether any more empty rows are there. Start keeping on moving the process. The same
way do empty columns which will identify all the last activities keep on identifying the
last but before activities and finish the process.
So, in between you have some loops and you have something with the present in the
loops. There maybe one or two blocks or there may be independent activities. And
dependent activities coming with that also you have to identify that. There are so many
methods of doing it. Power of adjacency matrix also you can use. Or you can use
primarily the path searching.
I have done path searching method what I have done is, graphical view of this complete
active matrice I have taken and I have shown it up here. And then I merged all the
activities which has the cycle length of 2. And then I have done topological sorting again
in order to understand the sequence. And then I have replaced it in my partitioning DSM
and I have shown, what is the sequence for the whole project ok. Now this is step
number 2 on DSM operations number 2 ok.
Now, once we have identified or done with the partitioning process, you should know
what is a sequence within the block ok. Should I do A first and then E or should I do E
first and then A or should I do C first and then D or should I do D first and then C for
which you may have to do tearing process.
So, the next operation in DSM is primarily called tearing process. Again I have taken a
DSM I have also shown as a different example. You can use all these examples for
practice also. This is one example wherein I have one start activity E and one activity in
the end as activity C and in between I have a complete block. That is one block which I
have ok.
Now, if I start evaluating this. So, what will happen is, B gives information to A ok, then
D gives information to A; then F also gives information to A ok. And then F gives
information to B ok, then A gives information to F then G gives information to D and
then F gives information to G ok. So, I have identified all the relationships A to F. So,
this is little clumsy. So, I am going to put it in another format A to F, then B to A ok,
then next is D to A ok, then F to A ok, then F to B, F to B.
Then next is F to G then G to D ok. So, this is primarily the graphical representation of
the block inside ok. If you see here what are all the loops I have. I have AF which is one
loop of cycle length 2. Then I have BAF is another loop which is of cycle length 3; B to
A, A to F and F to B this is a length of cycle length 3. A to F, F to A is a cycle of length
two. Then I have one more cycle which is DAFGD ok. So, this is primarily a cycle of
length 4.
So, I have three cycles of different size within the same block ok. So, now, what are the
options I have? So, I have. So, many options now, I am going to put this here. So, that is
easy for us to evaluate A F, B A F and then D A F G let me erase all these ok.
So, now what I have done is, I wanted to tear off the marks. So, what I have done. The
one mark which I wanted to tear is A to F ok, then B to A, D to A, D to A then F to A
then F to B then F to G and then G to D ok. So, now what I have done is I am going to
tear off this mark which is A to F. So, let me use another color. So, A to F so, this one
sorry this one is what I am going to tear ok.
So, now, if I have used A to F as my tear marks then my resulting sequence for execution
will be F and then G then B D and A. So, this is one way of evaluation because when I
have opened up the whole graphical representation, I saw this one relationship A to F is
actually you know causing three loops to be existing there.
If I break this one loop from A to F automatically all the three loops are getting are
getting erased away ok. So, I have chosen one with the help of my graphical evaluations.
There are other methods also available for the same. Let us assume I am going to use
strength of dependency. The weightage of the dependency between all these activities I
have taken up. I got values from my experts and I have dumped in all my values on a
scale of 0 to 1, which implies 0 means no dependency. 0.1 means is very weak
dependency which can be ignored, 0.2 a very weak dependency which may be
marginally ignored ok. 0.5 is like moderate dependency 0.9 or 1 is like a very strong
dependency. And if you start breaking it; obviously, you are going to have so many
repetitions going on. Now, blindly I am going to go with the NDSM this is primarily call
NDSM which is nothing but numerical design structure matrix ok.
Primarily I have got all the values within my blocked, because in tearing you are going to
find the sequence within a particular block. So, I have got all the values within my block
ok. And what I have done is, I know these are all the loops which I have. So, I am going
to break the weak links ok, what I have done is I have dumped in the values of all these
relationships over there.
And I am going to break the weakest link ok; let me write down again ok. So now, this F
to A ok. So, this is primarily my F to A this is 0.1 and G to D, G to D this is primarily
my 0.2; then F to B this is primarily my 0.3 ok. I have given only the weak dependencies
ok, let us try and see whether we can break the weak dependencies.
I am going to break this relationship, this relationship and this relationship ok, 1 this is
number 2 and this is number 3; because they are going to be my very weak dependencies
here ok. So, when I have thought of breaking all these relationships; obviously, I have
broken down all the three cycles in my loops ok.
So, this cycle, this cycle, and the third cycle. All the three cycles are broken down. So,
now, my resulting matrix will have a different sequential together. But I have to make
three assumptions. Now what is the variation between option number 1 and option
number 2? In option number 2, let us see here A to F. So, this is primarily a dependency
of value 0.5 and these are all very weak dependencies 0.1, 0.3, 0.2.
But I am going to make three assumptions and I am going to break all my weak
dependencies. Or I am breaking one medium dependency and I am going to use this as
my sequence or breaking three dependencies and they are going to use this as my
sequence. So, now, which is the better sequence among the two? That you will not be
able to see just with the tearing process because, you have to evaluate with the help of an
impact.
The measure of performance for us is primarily on time scale. Primarily you have to see
what is the duration for this sequence and what is the duration for this sequence; based
on the repetition that you have got all these values. How much time the activity is going
to repeat what is the duration and the reach repetition? And how are you going to see the
whole project that you may have to consider now.
For that we have to do the iteration modeling which is the last step ok. If I am going to
now take the same example which I have shown here, I have broken down a moderate
dependency. The same example I have taken up here and I am going to evaluate this. So,
now, what we will you do? So, I have also dumped in the duration along the diagonals,
diagonals always carry duration and they do not have any other value.
E’s duration is 4; F duration is 3, G is duration as 6, B is 7, D is 9, A is 5 and C is 8 ok.

Now, if I going to plot this on a bar chart. So, I am going to use a bar chart now ok, 0, 5
this is 10 this is 15 this is 20 this is 25 this is 30 and these are my list of activities. If I
keep it in that order, now this is the sequence that you have already identify. So, activity
E is taking a duration of 4.
So, this is E and F is actually a parallel activity to E; because I do not have a relationship
with E. So, I can start F even before here. So, this takes a duration of 3. So, after F, I
have to do my G and activity B and both are actually dependent on F only and they are
both parallel activities. So, I am going to do my activity G and my activity B ok.
So, primarily 3 plus 6 is 9. So, 3 plus 6 is somewhere 9 and the 3 plus 7 is somewhere
close to 10. Now so, these two a parallel activities ok. The next activity B and D are
again parallel for example, activity B is dependent on. So, after activity B, I have to do
my activity D. So, now, activity D, I am getting inputs for activity D only from G. So,
after G, I can start my activity D and it takes 9 days of duration ok. So, 9 plus 7 is
somewhere around 16. So, this is primarily my activity D which is stopping at 16 ok.
Now what about activity A? Activity A needs input till D. So, I have to start after activity
D and this is 16 plus 5 will be somewhere close to 21 ok.
So, now activity C can start anywhere before that. It is primarily supposed to start after G
after activity F and G are done I can start my activity C; but what about the iterations in
the sense what about the result of this assumption that you have made. So, once you have
done with all these activities. So, primarily after activity A is also done, then you may
have to repeat activity F ok; now without knowing the duration of activity F you cannot
do.
Now, let us make some rough assumptions. Every time the repetition is happening, one
third of earlier duration is taken for all those activities. So, what will happen here. F will
be done for 1 day in the repetition cycle. This will be done for 1 day this will be done for
2 days this will be done for 2.3 something this will be done for 3 days. This will be done
for 1.67 days like that it starts repeating ok.
And maybe if you want to stop the repetitions in the next round and so on, then primarily
that will be after one round of F G B D A F D till A. And then activity C you will get the
duration of the complete project. So, that is primarily how you evaluate. May be you will
get a duration as something as X. Now for this sequence we got an alternate sequence by
doing three assumptions here. There are three assumptions here.
Now, again you may have to evaluate this maybe similar kind of assumptions as to
duration of the activities are actually half of the earlier duration put it in that way. And
then repeat the activities may be one round or something as planned you earlier. So, take
up the duration maybe you are getting a duration as y ok.
So, primarily what happens is. You can do this combination for if you are doing literally
the iteration modeling, you can also skip the tearing process. And you can do this
iteration modeling immediately after your partitioned DSM is done that is also an option
of doing it ok. You can evaluate all your how many activities are there? 5 activities are
there.
So, 5 factorial combinations you will get ok. So, all this five factorial combinations with
the different sequences you can evaluate. And you will still get the minimum duration as
your preferred sequence ok. That is also one way of doing up your scheduling on
information modeling.
So, now, let me recap again whatever I have told you. There are 4 operations in DSM
formation of DSM, partitioning process, tearing process, an iteration modeling ok. So,
formation of DSM primarily there are methods for beginners, there are methods for
people who have come with experience and they can easily form just like an expert can
easily put up the network on CPM or PDM.
Same if we have very comfortable you can directly draw the DSM first go itself. Or I
have shown you method; use entity hierarchy either as teams, components, deliverables
parameters and so on. The entity hierarchy you can still use break down the whole list of
your project and then stop at wherever you want. Suppose if you are at conceptual stage
you can stop at teams component. Suppose if you have a detailed (Refer Time: 35:15)
stopped activities parameters are deliverables and parameters. You can stop according to
a comfort level and you can still form whatever type of DSM you want ok.
The next step is partitioning, partitioning is primarily because the earlier DSM that you
have done. You do not know, what is a preferred sequence of execution; we have
dumbed in all the data here and there.
In order to know what is a real sequence of execution, we have to reach sequence of

rows and columns. So, first step in there is primarily identifying all the sequential
activities: independent and dependent activities. And more them to the either of the
corners then you will have a large mass of activities which have some loops ok. There
can be single loop, there can be multiple loops. There can be multiple loops which are
over lapping with each other and forming a bigger looped. Any combination is there, so
for that simplest way to do is, evaluate them graphically like how I shown you here
which we call it as a path searching method. And start identifying the loops do the
sequencing here itself and dumb the result on to the matrix.
If you are working on excel sheets and you have a computer with you, then you can use
matrix multiplication. So, raise it to the powers of n and whenever you are landing up in
excel sheets which have rows, which are having values in the rows in the diagonal cells.
Then chose them as cycles or loops and merge them accordingly and you can proceed as
per the practice.
So, path searching or power of adjacency matrix, any method will help you to identify
the loops ok. Now, you know where are all the blocks and the cycles and you know you
will also know what is a real sequence of execution. The next step is, what is a sequence
of executing within a block for that you may have to do tearing process. And then you
we primarily selected tear mark again you can evaluate graphically or you can use a
numerical assessments also. Giving weightages to the relationships and you can still go
ahead and evaluating ok.
But, still you do not know what is the impact of your assumptions made. For that, you
can do iteration modeling. One thing is whatever options you wanted to try, only for that
alone you can evaluate. Or if you want skip the tearing process from partitioning process
directly jump to the iteration modeling step.
And still you can evaluate all the options available and still you can choose the preferred
sequence of execution. So, both the methods are really available. And as if I have told
you, what are the functions and the what is the characteristic feature of each of the steps
in the DSM process.
So, accordingly you can choose one or two operations conveniently. And you can still
get the scheduling of your whole project ok. So, this example will help you to understand
I guess. And many problems are given for practice and assignments still, we will come
back on doubts and we will clarify it later.
Thank you.
Lecture - 10
Dependency Structure Matrix-III
So, now coming back to phase four or the operation number four, iteration modeling in
simulation; so, simple steps again our given here for simulation exercise. So, first start;
select the blocking DSM for the iteration modeling ok. So, you have to really select. So,
so select the block in the DSM for iteration modeling. So, I have chosen one particular
block and for example, in the last example I have only one block. So, we can choose that
and decide the sequence of execution within the iteration block to initiate the iteration
ok.
So you should decide whether it is the ACEF, whether it is CAEF, whether it is CFAE,
whatever it is, decide the sequence of execution, which implies tearing has to be done ok.
Then substitute the X marks in the iteration block with the rework probability values. In
the example I have shown. So, there are so many X marks there, you should know what
is the probability with which for choosing an assumption and how much of rework will
happen in those activities and successor activities as well ok; that you may have to
decide ok.
I may make an assumption for initiating an activity called A ok. A is actually giving
information to maybe two other activities B and C. So, as a result of first you have to
check by making an assumption is A repeating or not and how much is A repeating ok.
And if A is repeating whether B and C are repeating or not and how much is B and C
repeating ok; this B and C maybe giving information to two or three more activities. As a
result of B repetition, is other activities are repeating or not? And how much is a
repetition?, how many number of times it can repeat and so on. That for that you may
have to get rework probability values.
This probability values you cannot do randomly you have to go to with the experts.
People, the domain experts, people who have worked on a project for last three times or
four times may be very comfortable in giving the rework probability values ok. For
example, the same cooling tower example I can say the GA arrangement we generally
takes 5-6 times. So, they can give a high rework probability for that and compared to
other mechanical drawings or deliverables that they have ok. So, the only the people who
worked in the projects earlier. For new projects it is very difficult to give the rework
probability values ok.
For projects which we have done earlier easily you can assign the rework probability
values and you can still get the values ok. Then once you have got all once you have
replaced all the X marks with the rework probability values and the next step is to do the
iteration modeling. Iteration modeling the best method is simulation, but that is not the
only one method. You can choose any method of your own choice; Markov chain is
there, Signposting is there, so many methods are available you can choose any of them.
In this exercise, I have used simulation and I have used manual way of showing with a
simulation exercise. And perform iteration model using simulation to obtain the iteration
duration of the block ok. Once this is done update the DSM with the new duration and
then stop the whole exercise ok.
Now, what is the meaning of block execution? So, I am going to take a hypothetical
example and then show you what is the real meaning of blocks? So far you have seen
what is a block? You know what is tearing process. Tearing you are doing only to
determine the sequence within the block ok. If you want to really evaluate the sequence
suppose I have 2 by 2 activities, a 2 by 2 block, then I may have 2 sequences. Suppose I
have 3 by 3 activity in a block, then I may have 6 sequences.
So, accordingly you may have to really see: what is the meaning of actually evaluating
the blocks ok. So, I having a block I have taken a very hypothetical example; A B is two
activities, duration for A is 12, duration of B is 8 ok. And I have used 0 as the
relationship between the two activities. 0 implies there is no relationship between A and
B so, I have put 0. Suppose if you if you want to have a minimal value in the duration,
then you may have to put 0.1 or something if you want to show the repetition. Otherwise
I have to keep it as 0, 0 implies no relationship at all ok.
So, now, what happens since there is no relationship, activity A and activity B are to be
done in parallel ok. Since there is no rework also shown here so, and what is the duration
for this particular block activity A, duration 12, activity B, duration 8. I have used bar
chart for showing this bar chart I will explain you in the end along with the other
techniques ok. So, duration for this particular block is coming out as 12 year ok. Now I
will take another case.
Now I should explain you what is rework probability now, with this example I will show
you what is the rework probability. Now same example I have taken, I have used a value
called 1 here. So, whenever so, this X implies X originally it is X, so it implies there is a
relationship from A to B. I am replacing this X value with 1 using a rework probability
value of 1. So, now, what is a rework probability value? As a result of change in A there
is 1 which implies 100 percent probability that B will repeat ok.
I am repeating it again as a result of any change you make in A, 100 percent of the times
B will repeat which implies every time you make some change in A, B definitely will
repeat because I am having a 100 percent probability value ok. Now if you look at this 0
value here, as a result of any change in B, A will never repeat, because I have given a 0
percent probability value ok. So, since I am giving 0, it implies it never repeats, 1 implies
100 percent repeats. Suppose if I have given 0.1 as my value as a result of any change in
B there is a 10 percent probability that A repeats ok. Suppose if it is 0.2 as a result of
change in B there is 20 percent probability that A will repeat. That is how it has to be
interpreted ok.
So, now, let us go through with this ok. So, now I am having an activity A, activity B
here. Now the duration for A is 12, duration for B is 8 and I have given a rework
probability of 1 here and 0 here. So, how does this actually evaluate. So now since there
is an X mark here so, activity A first and then activity B because there is a dependency
relationship. So, only after A, B executes and now what happens there is a 0 value here.
So, after B what happens A never repeats, because I have given 0. So, what happens is
the duration stops. So, if giving this as a block so, this shows A and B are in dependent
activities ok. So, in a way this is my parallel execution, this is my representation to show
my sequential execution.
Now, I am going to change the scenario. Now what I have done is I have written 0 here
and I have written 1 here. This implies A and B are parallel, because there is no
relationship between the A and B and this 1 implies every time B is changing, there is a
100 percent chance that A repeats ok. If you are just executing this model what happens
is every time you will get the same duration because there is no probability at all.
Here A’s duration is fixed 12 deterministic value, B’s duration is also deterministic given
as 8 and every time B repeats, you have already told A repeats and nothing is given on
information on repeat duration for A. So, what happen? So, 12 A it executes, then B also
is executed in parallel. So, after B, I am executing A in parallel and duration is 12. Now
what you have to really see is I am already not finished my activity A ok. So, should I
repeat my activity A to the fullest extent possible or can I overlap with the activity A
execution or should I wait for the activity A and then repeat my activity A.
So, all these queries that you have, with which you have to really carefully plot the
values and give control statements for your activities so, that the activities will perform
as per your desire ok. If you do not give any controls, default what happens this is what
is the execution that will happen ok. Suppose if you think since half of the execution is
already merged here ok. You need not to repeat this much of extent, then you can cut
down the duration of repeat, you can do something here ok. So, those are you may have
to make it in your assumptions case.
Now let us put another case ok. Now I have done 1 here and 1 here. So, what happens
here, 1 is like every time A repeats, B repeats. Every time B repeats, A also repeats. So,
what happened to this network? I have only showed two times it does not mean the
model is chopping at two. In this case what happens the model will never even stop ok,
because I am executing activity A, then activity B because it is sequential as a result of 1
here and as soon as B repeats, A also repeats, because I have told it is a it is 100 percent
probability. As soon as A repeats B also repeats, because I have told 100 percent
probability.
So, what happens A B keeps on repeating every now and then ok. This is also a caution
you have to keep in mind; the model will not even stop. Whether you are working on
simulation model it will never even stop. So, what you should do; maximum you can
give 0.9 or 0.8 as a value for your rework probability values and be careful when you are
keeping a value as 1. Unless you know that 1 will not no stop your model so, do not keep
a value as 1 for example, in this case that may happen ok.
Now what are all the assumptions that one can make. One is duration as I told. When
activities are overlapping with the repetition in generally does not happen in information
flows. So, primarily the duration of the repeat activities need not be same as the initial
activities and other thing is so, suppose I wanted to cut down, maybe I am just planning
for half a duration is a duration for repeats ok. In the next query that can come is how
long can I repeat do the repetition.
In real world practice if you will see when you are modeling on information and
revisions in drawings, any sequence of drawings will not repeat like 10 times or 20 times
and so on. It hardly repeats 2 or 3 times. So, you can put a control statement ok. It should
not be less than maybe 1 day of duration or it should not be less than 2-3 times often,
then you should the model should stop in the repetition. You can give all these control
statements when you are doing your modeling ok, then only you will get the duration
which is really realistic ok.
So, I am just going ahead with the same example. So, what is the meaning of the block
execution in the design phase ok. We are still working on design phase so; we have to
really see what is a meaning of the block? In the last example case I have shown you:
what is the meaning of plotting the numbers ok. So, each number and different numbers
and different places, duration is still the same. So, I have told you: what is the meaning
of the same. So, what are the assumptions minimum assumptions one should keep in
mind and in order to get a scheduling duration and also what are the things you should
understand in the block execution ok.
And so, now I am just evaluating it with the last slide you understood that in the
repetition will keep progressing till end, there is no limit ok. This is a meaning of this
whole block. Now what happens, A starts because I only made this as my assumption
because I have given the sequence as A and B it is understood there is an assumption
made for this link which is this link so, there is an assumption made in order to initiate
my execution.
So, A is starting first, next B, and then A repetition, B repetition first time, A repetition,
B repetition second time and this goes on till the end ok. That is understood. Now I am
just giving the same duration for the activities and now what happens is I am just going
to see what is the minimum and maximum in these cases which I have told you if there is
some value in the dependency relationships rather than 0 and 1 which are the extremes.
So, you will get a duration minimum as 12 and 8 because of the sequential relationship
and the maximum will be so much ok.
So now, there are some assumptions to be made. So, number one assumption what I have
done is: the repetitions of the activities, I have said, it takes half of the original duration
only and stop or 3 rounds of the duration. If that is a case, then this will be my duration.
So, this is hardly 10,10, 20 and this is 20 so, 40 is a duration of my entire block ok. If
this is assumption you have given in your model and you are doing any probability
value. You are giving a probability value of X from anywhere between a 0 to 0.9 or
something and why also if you give from 0 to 0.9 something, what happens is your
model will oscillate between the value called 20 and the maximum value as 40 ok,
because I have made a control statement here ok.
So, that is why what happens is if you want to you can also put an interpolation or
something and you can get the intermediate values manually if you are not comfortable
in doing it on simulation ok. So, now, what are the assumptions you have to do, really if
you want to do this also the second repetitions also you can say it is half of the earlier
repetitions. So, in this case it can be 2 it can be 3 and 2. Next round it can be 1.5 and this
can be 1 and you can say values less than 1 are not even considered. So, what happens
this will stop at here, because half of 1.5 will be 0.75 and you cannot go beyond that.
So, the model will stop on that place. You can give assumptions in whatever ways you
form according to the model representations ok.
Now I have shown for different XY plots, but there is no control on duration here that is
why I got values more than 55 and all. Otherwise you will get between that values only
ok. In this case I have not shown anything ok. So, if the value of X is 0. So, the value of
X, I am giving it as 0 and the value of Y I have constantly kept as 1, which implies that
there is a strong relationship between A and B and that is where I have not removed that
relationship and the B to A is actually a relationship which can be removed or ignored
ok.
So, I have maintained this as 1. So, what will happen is every time A repeats and then B
repeats, if the value of X is 0, then it will stop at 20 only. The value of X is 0.2, then it
repeats for a few rounds ok. The meaning of that is suppose I am doing 100 runs in
simulation, 20 percent of the time A will repeat and remaining 80 percent of the time 80
runs ok, in 20 runs A will repeat and duration will be a 6 or something like what you told
and in the remaining 80 rounds what will happen is A will not repeat ok. That is the
meaning of that.
If you are running like that, you are duration is something as 24, if you are changing this
value as 0.4 it is 27; so, with the increase in probability value so, you can understand
something. What is a meaning of this increasing in duration? It means actually the value,
the relationship between B and A, it’s strength is defined with the duration ok. You have
made a very strong assumption when I am saying there is a 0.4 probability it implies
whenever there is a change in B, 40 percent of the time A will repeat which implies the
strength is really greater than 0.2 or 0 and something.
So obviously, the repetitions is expected to happen ok. You should understand when I
breaking a tear mark with 0.8 or something, we should understand and my durations will
keep on repeat every now and then. And, I may get something close to 50 60 or
something for my initial durations on 12 and 8 ok; that you have to understand and keep
in mind ok; so, as far as possible try to avoid breaking up a mark which has a very high
and value in the rework probability values.
Now same case I have again shown wherein both the relationships are changing equally
ok. So, when X is 0, Y is 0. So, it implies the relationship is parallel. Sorry for this
change. So, this will be now 12 ok, then X is 0.2 and Y is 0.2. This implies the first
relationship is still there and then there is a rework probability of 0.2 and 0.2. So, this
will be 21.1 then 0.4, 0.4. So, accordingly you also see here when the strength of the
dependency relationships are increasing, you also get an increasing pattern ok.
Now if you see let me explain you the pattern. So this if it is this is greater like 1 or
whatever it is, you will not have that much of an impact ok. This X mark relationship
should be very less then you will only have little rework. Suppose if this relationship Y is
more compared to X, you will get a greater duration than X less and Y more ok. That
you have to understand and if you are having both the values as close to 1 or something
what will happen is the model will not even stop. It keeps on running because every time
both the activities will keep repeating ok. So; obviously, you have to have a control
statement when you are working on information flows.
Now so, this is all about on information, again after explaining simulation I will tell you
how do we run, how do we plot the simulation model as again working on computer
simulation modeling and helping you with the software is not part of the course, but at
least I will tell you how to do it with manually on the simulation models which I will
postponed little later. Now let us see the same cycle flows, let us see how to represent in
a DSM, so representing workflow in DSM.
Suppose if I want to use the DSM example in a construction phase. So, what happens
and what are the changes ok. So, the researchers have identified that there are two ways
of executing a sequence ok. One is one is called the basic sequence and the other one is
called the execution sequence ok. Now basic sequence which is nothing, but C after that
A and E are in parallel. Then F activity and B activity, because this is in parallel and then
activity D is executed. This is a basic sequence. This you will get from where? As soon
as you are finishing a tearing process you will understand that this is the basic sequence
for execution and for this entire project ok.
Now, what is called execution sequence? Execution sequence is what is the order with
which I am going to first execute my CAEF and repeat my CAEF ok. That you will
understand only when you are working out your execution sequences ok. There are so,
many ways of doing it.
Two methods are described here; one is slow and one is fast. So, what is a slow a
strategy implies? So, after C, I can start my A, but there is an option for me to repeat and
check my C also. But I do not want to do because I do not want to repeat activity C too
many times. I am going to wait for my activity F, information from F also let it come, let
me cross check my A and F and then I will repeat C accordingly. That is also possible
ok. So, what I do; I start executing my A, then executing my E in parallel, because there
is no dependency relationship here then F. Once I have done, then the second time
repetition of the entire block happens, because it is a slow process of execution. Slow
implies I do not want to repeat too many times every now and then as soon as new
information comes I wait for all the information comes from all the activity execution
then I will start my repetitions ok. Like this the repetition 1 starts, same way repetition 2
will start and so, on ok.
Now what is my fast strategy. So, as it is shown here, now once the activities are done
then I do my activity B and D execution in the end. That is what is a slow. In fast what
happens, I start repeating my C as soon as my activity A is done and once my activities
within the block are done first time itself I start repeating my activity B D also. So,
outside the block also I start repeating and as a result of each and every time of an
execution what happens is I keep repeating my activities outside the block as well ok.
So, this is an information flow mechanism ok, which implies and as an information
arrives, you keep executing, and also passing on the information to every other team.
And then so that if they are doing those activities they can also cross check this value and
then the errors are assumptions can be made minimal. In this case repetitions can be
more if you are working on a design phase you may have more versions on the drawings,
you may have more versions in the design mechanisms, but still you may have a shorter
duration for the entire design phase.
That may also happen ok, but let me put the same concept, slow fast concept in a
construction flow ok. Let us see how it works ok.
So, I am having a earth moving operation, this slide I think you remember this. This I
covered in the first and second class of the lecture ok. I am having an earth moving
operation. So, this is my cycle flow. So, you know what is this example all about. Now
where do I have the cycle if I want to know only the activities if I am going to model
only the activities here, how many activities are there loading, hauling, dumping and
return. There are four activities available in this particular example and one thing you
should know is all these four activities are in cycles now ok.

Suppose if I want to model this, what happens is loading, hauling, dumping and return; I
gave some rough duration for all these based on some assumptions on data that they
collected from the site. So, it takes it takes 2 hours of loading operation, 6 hours of
hauling so, primarily traveling on the road, it takes 1 hour to dump the soil and it takes
same 6 hours return, because I am going to travel on the same road only ok.
So, I have made the assumptions. So, 2, 6, 1, 6 and a what is a cycle flow now. After
loading, I have to do my hauling operation, after hauling I have to do my dumping
operation, after dumping I have to do my return process and once return I have to again
go back to do the initial loading operation. Now what happens here this is a very good
example slow strategy, but if you see here this looks as if I have one resource with me if
I have only one dump truck with me, then I may have to do the cycling only after 1-1 full
cycle is completed ok.
So, this is also a resource a pattern which is also revealed in this example ok. Slow
strategy if I have only one resource in workflow this is what is the meaning of the slow
strategy. So, I have to execute all the four activities in one round, then I have to repeat all
the activities in the second round and so on ok. So, roughly takes 15 days to finish my
one cycle, then another 15 hours sorry 15 hours to finish my second cycle.
So, hardly it takes more than a day for me to finish if we are working on 8 hours pattern,
it primarily takes one full day, I would say less than a three-fourth of a day or if you
want to stop and then continue on the next day after 8 hours then it takes more than 1 day
for calculations ok. So, that is what is the meaning of this. Now what is a meaning of fast
strategy? So, if you want to add, if you want to start your loading operation let us
assume, I may have more dump trucks ok. So, what happens is I am going to you know
fast track the mechanism ok. What I am going to do I have introduced one more X mark
here ok.
So, this implies so, as soon as my return is done, again I do my loading as soon as my
hauling is done also, I can still start my loading. So, how will this be executed. So, load
to return processes is coming till here, it takes the same 15 days ok, but as soon as my
hauling process is done I can still start my loading, that is also done and as soon as my
return operation is done still I do my one more loading also ok.
So, after this return I can do my loading also. So, this is like multiple sources are there,
you can still go ahead with your loading operations and so, on ok. But in real world
situation what happens is as soon as the first dump truck is loaded up and sent, we do not
wait for the hauling itself to take place ok. We start loading it up as soon as the first
loading is done which is something like this ok. If you want to work something like this
use the same fast track strategy and add controls on the statement so that the activities
keep repeating every now and then ok. So, this is the meaning of using the DSM in
construction phase ok.
Now as I told you earlier DSM is an ideal method for information flow in design phase
as well as in construction phase, but I can also use the same DSM for workflow in
construction phase as well ok, which is what I showed you with this example. You can
still show multiple resources and use execution strategies for showing your multiple
resources. So far what we have seen is DSM, even in the same DSM when I am having a
block for example, this block which I showed you this also you can use it in information
flow, but I showed an application in construction because you should not have the
mindset that it is only applicable for the design phase ok.
So, you can go with the slow strategy you can go with the fast strategy also, but in the
AB example there is no meaning on slow or fast and both the values are same. Why,
because there is nothing called waiting. The minute you have more than one X mark or
one feedback for any activity, then only the slow fast strategy generally happens. For
example, in this case if I have one X mark here and one X mark here, the slow or the fast
will not be applicable. Both the values will be the same for you ok.
Now, let us move on into a what simulation, because I have told you so far primarily the
process behind simulation. And so, far I have told you only the manual way of showing
the simulation. Now if you are having a very large models in this case you did not have a
very big model. So, you had only a smaller example. So, we were able to easily show the
manual executions ok. Now you have to understand how to do this computer simulation
modeling also. So, we will see what is the simulation process as such ok. Little glimpse
on what is simulation, notations behind that, and so on we will see it right now.
What is simulation? So, this is a continuation on iteration modeling. You can still go
ahead with using interpolation and you can still have a rough value on iteration modeling
manual way, Simulation if you are very comfortable any of the software then you can
still go ahead and simulation in getting that real simulation values ok. So, it is an
imitation of the operation of a real-world process or system over time. That is a
definition given by researchers on what is simulation. So, what is a simulation model? It
is a particular type of a mathematical model. So, in mathematical models you have
analytical type and numerical type. So, this is a numerical model where in the models are
run every time rather than solved. So, this is a very it is supposed to be a very useful tool
for planning and decision making also ok.
Now whenever you say simulation model, primarily people say it is a computer
simulation model and associate the word called computer with the simulation. This
implies since real world simulations are very large with lot of data analysis and inputs,
the runs are not possible with the help of a computer. For example, even in the previous
example, you cannot do this manually ok. Suppose 20,000 meter cube of earth has to be
cut and every time 10 meter cube of earth can loaded up on to your dump truck.
If you want to calculate this, then it is very difficult to do manual calculations unless you
want to do some analysis like this and do ok. The best way to do is and maybe let us
assume and sometimes in the peak hours of travel there may be a traffic jam so, you may
have two hours of delay; non-peak hours you are actually able to come fast. So, if you
want to plot all these uncertainties and durations and so on, you cannot do a manual
calculation with the computer or using a calculator. So, you need a simulation modeling
for the same.
So, the simulation always it goes along with the computer and we always say it is a
computer simulation model ok, but still simpler models can be simulated manually. In
the previous example all I showed you the manual way of explicitly representing a bar
chart ok. Now what are the advantages of simulation, because I introduced you use
simulation you should only little more on what is simulation.
Advantages; it allows decision making. The primary advantage of going ahead with the
simulation model is, as an experimental technique is, it allows decision making because
you cannot really make a trucks move on the road and then calculate and do something.
You have to really take decision making on how many trucks I need, how many loaders I
need, how many excavators I need, those decision on how much time equipment’s are
waiting in the spots, how much time the resources are taking, then can I cut down the
duration here and so on. Those type of decision all if you wanted to make then;
obviously, you have to go in for a simulation model.
A simulation experiment can run any number of time, this is an advantage as an

experiment ok. if you are not comfortable change some parameters run the model again
ok. If you are still not happy with the result ok, modify all some parameters, change the
network run again experiment how many of a times you want ok. The experimental
results are just obtain a few minutes or seconds rather than once the model is really
created, rather than you doing it manually on something ok.
The sensitivity of a model to changes in inputs also can be tested. Suppose if I want to do
a breakdown along the path, 10 times the truck goes and there is no problem the eleventh
time it travels on the road that is a breakdown and then 2 3 hours of delay happens for
repair and so on. If you want to model or what is a sensitivity if I calculate my duration
as 10 days compared to 11 days or 7 days and so on. All these also you can really test
when you are running the model ok. Sorry for the two disadvantages here it is a mistake
ok.
Disadvantages; so, the modeling requires special training and as I told you each software
is really different. When compared to even the software the basic representation of
simulation itself we have to really understand so many control statements, constraints,
loops and so on. You have to understand that properly when you wanted to do and you
may have to do trace runs to see how the model is really performing, as per your plan or
is it going in a different way you have to do all those. So, it needs special training.
Even representing or interpreting the simulation results, that is also really difficult and it
takes so much time for interpretation. So, simulation modeling and analysis, both are
really time consuming and expensive based on the software. Sometimes you may have to
buy software and then ask for training, learn the software and then you may have to do
ok. Lot of software are available in construction some are free for research or academic
institutions and many of them are primarily commercial software ok, that you have to
understand.
The common terminology, in simulation without which you have to understand that:
number 1; random number generator. So, everywhere there is an uncertainty in a
simulation then that is because of the input on random number generator ok. Every
simulation package has a random number generator. In the last example on AB example
I gave values on duration as 12 8 and so on. There is no uncertainty there. Suppose if I
am giving a PERT distribution or I am giving a uniform distribution or I am giving a beta
distribution then; obviously, the random number generator starts playing a role there ok,
and in the example last example on earth moving example I gave so much of
uncertainties there.
The truck takes different time when its travelling on a peak time, non peak time it takes a
different time, there may be a breakdown during its travel sometimes it happens. So,
these are all randomness or uncertainty in the whole processes which is generated with
the help of a random number generator ok. Every simulation package has this. The
random number generator is primarily called a pseudo random number generator. It is a
software routine that generates a RN between 0 and 1 that is used in sampling random
distributions.
If you are very keen in doing this open excel sheet and type random and then you press it
down ok, you well get a different value on the complete sets ok. You will you can do it
later and check. The next thing which we were understand in simulation is determination
of simulation runs ok. If you are really done the simulation only for one time then
obviously, you are not going to evaluate all the parts. For example, in the AB example on
probability as I said 20 percent of the time A takes repetition and 80 percent of the time
A does not repeat.
If you want to really get an average of the entire runs, then you may have to do minimum
100 runs or maybe even 1000 runs in order to have a average which is close to the real
number ok. If you do just 1 run or 2 run whatever is a randomness created, you will get a
different value which may not be closed to the real execution ok. So, for example so that
is primarily the number of simulation runs you want to carry out in the experiment and
do an average of 50, 100 or 5000 runs and then you get the value on the duration for the
project in this particular case ok.
Sometimes what happens is if the time the problem is really-really very big, then the
time taken for 1 round itself may take maybe 2 hours or something. Suppose if I want to
do 1000 runs or maybe even 100 rounds it is not practically possible ok. So, investigation
itself has to be determine as to what should be the number of simulation runs and with
the minimal run will I be able to get a result which I can interpret properly or not that
you have to really analyze ok. Sometimes even pilot studies are really done in some
cases to really examine what is the number of simulation run one should ask use in the
particular model ok. That is also some cases happens.
The next is progression of time ok. There are two clocks in most of the models. One is
called the clock driven and the other one is even driven. Clock driven is like hours like
10 am 11 am 12 am and so on and the other one is event driven ok, which happens as
with the case and a primarily it is like a duration takes 2 hours so, it shows first 1 hour it
shows a first 0 and the 2 hours and then after 2 it jumps to 4 and so on ok, which is an
event driven which happens ok. Most of the software are event driven rather than the
clock driven which also you should keep in mind when you are running the models.
Now, a simple step by step approach when you are using a simulation model ok. Number
one step; so, define the problem, establish the boundaries and the components of the
system to be modeled ok. In the same last earth cutting example I am just defining the
problem. The problem I wanted to know what is a cycle time it takes for excavating
some large bank of sand ok. That is primarily the problem which I have defined. Suppose
if you want to start from no finishing of the earth, then tamping, then finishing, clearing
of the sands, if you want to expand the problem as to other components and all then you
may have to properly choose your boundaries ok.
Then so, first define the problem earth cutting example. So, take only the problem and
then finish up all your components. I have only four activities which I wanted to do. I am
only keeping all those activities. I do not want to really see where are my trucks, when
are they coming in, where are the workers, how are they coming in. So, you do not want
to do other boundaries and all then you should trim your problem accordingly and
choose only the problem what you wanted to model ok.
The next; develop the logic and flow diagrams. In the same earth cutting example I told
do what is a logic diagram and logic diagram is primarily the pictorial representation of
the model that you have to do and you may also have to show understand what is a flow
process. The flow of the earth, the cycle moment of your trucks, cycle moment of your
loaders, excavators everything you may have to understand in the logic and flow
diagrams ok.
The next, conceptualize the model in ACD. ACD is nothing, but activity cycle diagram.
This we will explain in the next slide. So, you have to now convert this logic diagram
into an activity cycle diagram ok. The next step is data collection, data collection is how
much time the loader takes, how much time, what is a capacity of the dump truck ok,
how much time the dump truck takes to hauled, then where is the dump dumping area,
how much time it takes, should there be a queue there for the dumping process, earlier
dumping has not been done then you should wait or can you dump parallelly. So, all
these data on the model all you have to collect it ok.
Then transfer the logic network into a simulation network. This is where your software
parts comes into picture. Do you want to use software x y or z which plays a major role
here. We have so, many software available in for construction simulation especially apart
from genetic simulation softwares ok. Starting from your UM-Cyclone, Strobel strobe,
easy strobe, we have a Symphony, we have ExtendSim, we have a analogic. So, like this
so many software are out there. So, choose anyone which is comfortable for you and
each one has different notations and they have different ways to the model, but common
terminology all I will explained in this lecture itself ok.
Then transferred it there, run the model for getting the results, validate the model, see
whether the model is giving a result properly and are you able to get the result of your of
your investigation steps. If the model is giving a different result than what you are
expecting, then you have to redefine the boundary of your problem and then see whether
you are getting. I am not talking about the expected result, I am talking about the
performance of the results ok. That you may have to understand, then you use the model
for decision making. This is a simple step which you may have to use in for simulation.
Now as I told you so, let us see what is this? This also I have shown you on ACD. This is
primarily called ACD diagram ok. This you I would have explained to you while I was
taking you on to earth cutting example in the first few classes. So, this is primarily I have
two nodes here; one is a circular shape and the other one is circle or oval and the other
one is a rectangle shape ok. So, ACD’s primarily a method to describe the interactions of
the objects in a system ok.
So, here primarily I need three resources; loader, soil and truck. So, accordingly I have
given three different resources, then loading is my first activity to take place ok. For
every loading activity I need three resources for me. So, which is primarily the passive
state of an entity or you it is also called as a resource, that is primarily called the queues
in the circles. So those are all represented in circle form.
The active state or the activities are primarily called active states and which is
represented in a rectangle shape. So, I have only two different shapes to represent my
ACD diagram. So, whenever you are defining your boundary of your problem, after
defining the boundary, transfer it in ACD forms like this, then your model is very clear.
Then it is easy for you to transfer into any computer simulation software that you are
very comfortable with ok.
Next is notations and simulation. So, each different software uses different shapes and
options for this, but I am just going to explain you the common symbols which is
available in all of the software ok. Number 1 is queue. So, all these circles in the passive
states are primarily represented in a queue form and these rectangles can be either a
combi or a normal activity.
For example, this load I have to represent it as a combi activity, because I need resources
to start the loading activity. Hauling can be a normal activities, I am going to use a
general rectangle. This is also a combi activity, but this is a normal activity because there
is no waiting required here and the other arrow marks are all called as the links ok. That
is what I have shown here. So, queue it is a named element that holds idle resources. So,
at the beginning of any simulation, queues can hold certain number of resources.
The next is combi or conditional activity. It is a named element that represents tasks that
can start whenever the resources available in the queues preceding it are sufficient
enough. Suppose if I need two resources in the predecessor queue, then only when the
two comes in then only this activity will start otherwise it will not start.
The next is normal activity or it is also called bound activity. It is a named element that
represents tasks that start whenever any preceding activity ends. So, as soon as any
previous activities stops, it just follows a finish start relationship and then it starts
executing the same. The next is link it is primarily a connector between all these queues
normal and combi activities ok.
Now this example we have seen earlier ok. I told you how the durations are all varying,
escalating with the varying in the combination on rework probabilities. Now, you know
what is the simulation? Now if you want to represent this same activity and example on a
simulation network, then how do you do ok. So, you know this is a DSM example.
Boundary of the DSM is fixed. I have only two activities. Duration for the two activities
are also there. Rework probability values are also given. Now, how do I represent this
ok, it is very simple.
So, I am going to start with activity A. I am going to use my simulation network only
here or I can also show with an ACD diagram. So, there is a queue and then I am going
to start my activity A. So, this is a combi activity, queue is primarily the assumption here
ok. I got the assumption as a result of executing B to A and then, I am going to start my
activity A and once I am starting my activity A, I am not going to wait for anything. So,
this is my activity B. Duration for A is 12 for B is 8 ok. It goes on.
Then I have to repeat so for that if you go back here what happens is, the durations and
all you may have to give control statements for first second and so on ok. What I am
going to do and when you are having 0’s here as a Y value sometimes the B does not
even execute ok. So, I do not want to take a risk now. So, I am going to take the
repetitions very separately ok. So, you have to run the model and see how the model
executes in order to execute the whole ok.
Now I am going to use a probability value here ok, which says whether my A repeats or
not. So I have finished here. This Y is represented here, in the form of X mark and then
now I am coming back to here primarily I am going to evaluate this ok. Now I am going
to have two options here. So, this option will be the real X value ok, it can be 0, it can be
1, it can be 0.2 whatever it is. So, what does this say? So, according to this activity A
repeats and now this probability link will be 1 minus X and here it goes to finish. So,
which implies A will not repeat. When the A does not repeat, what happens? the model
stops, model has to stop ok. So, I am going to take it to finish.
So, finish is a place where I am going to collect the duration. The maximum of duration
that comes to finish activity, will be the duration of the particular run which I have done
in the model, that is how the model works ok. Now, when the activity A is repeating then
again I am going to introduce a probability here and this time what happens this
probability is primarily Y. And, I am going to say it is primarily the repetition of B and
again I have one more case which is 1 minus Y as to the B will not repeat sorry erase it
ok, as to the B will not repeat ok.
Now what happens here is, whenever my activity A is repeating so, it just goes to B
repeat ok. So now, I am having a two case. So, B can repeat, or B can go to finish ok.
Now this cycle is not completed if I leave it like this. Now what I should do with this.
Repeat B is there and this repeat B can also B I have to connect it with this probability.
So, then the repetition goes on as per the X mark. So, that the model is tightly packed,
and I have to do: duration for repetition A, duration for repetition B. I have to really see
whether it is half of the earlier and so on. So, that I am having a decrease in duration
based on the assumptions, sometimes my assumptions are too bad and I have to repeat it
again. Leave the duration as it is and then you have to show the control statements ok.
So, I have also shown you for the same model how do I put my simulation networks
representation so that I am showing the representation on a simulation network. Write
the code or show it on a graphical GUI based, show control statements on repeat A,
repeat B on the four clicks then you will get your model done ok. With this only we have
got the data here like this for X whatever value I have shown. In this case let us assume
if I have used 0.2 here, then this was 0.8 and the same case if this is 0.2 so, this was 0.2
as a link and this was 0.8 as a link ok. So, that is all on DSM dependency structure
matrix. So, all the four phases we have explained in detail with an example ok.
So, next class we will see with another topic, bye.

Lecture – 11
Beeline Diagramming Method
Topic of today’s lecture is, Beeline Diagramming Method. So, what is Beeline
Diagramming Method? This is a very simple lecture compared to any other any other
topic we have seen so far. Now what is this topic primarily about? So, let us introduce
little on the background side first.
So far in information management, we have seen DSM.
DSM I have shown you as to, how to represent information; but primarily it was like
only one relationship between any two activities or any two deliverables and so on. You
would have seen only one relationship. And DSM can also be very powerfully used for
representing construction phase also, for representing the workflows in construction
projects as well you can still use that also, I have shown you how to do and so on ok.
Now, what is this BDM representation? As I given the hint right now. So, when you are
having multiple information flows between any two teams or any two components and so
on; then you should look at a technique like BDM, for representing and for analysis. So,
what happened? So, have we not dealt all these earlier, have we not come across multiple
information sharing earlier and so on, yes we did so.
Suppose for example, if you want to see the history with which we have come across,
this I have also shown when we were discussing on DSM, the introduction to DSM.
Earlier we were very comfortable in making assumptions, all the assumptions varying
the schedulers mind only; and the activities were represented and the people were very
happy to go with sequential and parallel relationships alone. Then came in techniques
like simulation, DSM, and so on, wherein people could easily define the sequence, do
schedule analysis with multiple options, constraints and etc. So, the cyclic
representations in a way, we call it a cycles or interdependencies they were we could
easily represent, this happened maybe since 2000 and so on.
So, BDM is a very recent concept, which has come only in 2012, 13 and so on. So, the
person who developed this BDM was Kim. So, BDM the abbreviation is Beeline
Diagramming Method and these lines are primarily called as Bee lines.
So, if you are having a problem which you have to represent like this in multiple
information relationships; then you have to really go for BDM only ok. Now what are
these? There are other terminologies also for representing this concept; one is called
overlaps you can see here A B C are primarily overlapping with each other. So in a way
we can call it as an overlap, but the real overlap as you see in BDM that is not the case.
So, this is like, I would not say an advanced overlap, but it is too much of information
relationships and typically you are doing an overlap.
And in the last class, in the previous 1 or 2 classes earlier I have introduced what is
called linkages. Linkages are nothing but links. They primarily need not follow the
conventional Finish-Start relationship. So, these are all called linkages, I can start
anywhere from my predecessor and I can stop anywhere at my successor.
So, for this particular linkage so A is predecessor and C is a successor; and like I can
start anywhere in the flow and I can stop anywhere. So, it is not following in the
conventional Finish-Start, nor it is following like PDM representations wherein you are
able to show Start-Start, Finish-Finish, Start-Finish and so on.
So, you can actually directly connects the predecessor and successor under relevant
points. So, links are primarily the shortest travel distance or shortest distance between
any two activities or any two deliverables or any two teams and so on. So, they are
primarily called linkages. So, this Beelines are the linkages ok, they are also called
Beelines and they are also called linkages.
If you want to give a proper definition for the linkages, then this is a proper definition
we have come across. So, one is multiple you can represent so many relationships or
information exchanges among any two activities. Then it also is a two-way, it is not like
I am only having all relationships only in one direction; I could actually represent all the
relationships between any two activities.
I am not having only one side of communication flows; I am actually having two-way
communication flows. A can also give information to B, B can also give information to
A. So, it is multiple, it is also two-way information exchanges. So, that’s where we are
actually in BDM ok. This technique as I told is a very recent origin and we have
developed and we found that it is a very useful technique for representing an information
flows in design phase.
And what is a BDM representation right now, having known that I can represent multiple
information flows and so on and it is a technique for scheduling in design phase; then
primarily you should know how are you representing the time phase of all the
relationships, because information is driving your schedules here.
So, let us assume I have two activities; activity A, activity B and the duration for A is
given as 10, duration for B is given as 12. So, I have these two activities and there are
three ways of representation as given by Kim; but we are only going to show only one
representation which is the N1-N2 ok.
N1 is primarily the time it takes to show my time from the predecessor what time the
information can be released. And N2 actually represents the time at which the
information is received by the successor; that is how the information is represented as a
linkage.
So, N1, N2 is a linkage and it shows the time at which. So, N1 from the start of A
activity what is the time it takes to release that linkage maybe I can call this linkage as
P1 ok, this gets a parameter. So, this is the time it takes for the parameter P1 to be
released from A N1 and the same parameter P1 is required by B at the time called N2 ok.
That’s how the BDM representation goes.
Now, how can I get information directly from the site for this? For the first project;
obviously, it is very difficult for you to get an appropriate information. I would still say
you will get information on different parameters or different deliverables and so on you
can still get it.
But if you want to really work, maybe on abstract level like components or teams, then
you can still draw a BDM ok. I may have so many teams, I may have different
deliverables coming in out of those teams; still I can form a BDM and I can still go
ahead with that. So, this example is primarily teams and parameters, I have shown you
teams and parameters. If you want to get information like this and work out with the
parameter level on the teams or components; then you should have done at least 1 or 2
projects earlier ok, then only with that domain expertise that you have, you will be able
to say what time what parameters can be released and so on.
Because N1, N2 you would not be able to define unless you have done 1 or 2 projects
here earlier. Especially when you are working in the lowest level of your WBS; abstract
levels of WBS yes, you can still do it for the first projects.
So, I am taking a highway design case ok; there are three teams here. This is a truncated
example it is not a complete example only to showcase the relationships that travels
across the three teams. Three teams are there. Duration for the three teams; this is taken
from the time the three teams were on the whole project; 240 days highway design team
was in the whole project, 156 day structures design team has spent the time in the whole
project, hydrology design team has spent almost like 50 days in the whole project.
Now, so first link 14-6, so this implies, this is my predecessor, so this time is primarily
14 and the time with which it is reaching the structures design team is 6. So, what is the
information that is traveled: alignment details, proposed by-passes and the cross section
width, ok; so these are the information which has reached the structures design team ok.
Now let us stay see, from hydrology to structures design team at the time 25-6. So, this is
primarily 25 and this is my 6. At the same time, structures design team is actually
receiving two inputs; one from highway design team and one from hydrology design
team. From the hydrology design team, high flood levels, low water levels and so on are
received from here ok. Now let us see, what other information are travelled. Once this
information is received by the structures team, the structures team starts working on their
design calculations and it can actually send structure improvement, embankment slopes
details to highway design team. So, time frame is 8 and this is now 21.
Now that is how you have to understand that, at the same time high hydrology design
team is releasing information; linear waterway and number and length of the spans to
structures design team. Then after some time structures design team is releasing vertical
clearance to the highway design team ok, then finished road level at 16-34. Then from
the highway design team, after getting all the inputs it releases highway cross sections to
structures design team. Then design discharge is released from hydrology design team to
structures design team. So, like this the flow continues and all other information between
these teams are really exchanged. I have not shown the complete list, this is only to show
what is a detail ok.
In this case all these parameters which were exchanged between the three teams.
And other case ok, let us see the other case which is a metro construction example. This
is a metro design example and this example I have introduced you to show the complex
information exchange between different teams. And in any project you may have so
many teams and there may be complex interactions between all these teams, that I have
introduced you at that time.
Now, I am going to take the same example right now. So, this team this particular
example 20 teams we have taken out and within a span of 3 years there was close
interaction with all the teams that was there. Now the three teams which I am going to
take for today’s class will be, for explaining the BDM representation will be contractor
team, structural design team, electrical and mechanical team ok
Now let us see the three teams, so I had given a short symbol for the three teams. So,
structural design team, electrical and mechanical team and there is a contractor. So, there
are three teams given here ok. And there are few deliverables as a part of the three teams.
So, the deliverables are listed out here; preliminary design and design reports, definitive
design and design reports, construction reference drawings for cutouts. These are the
three deliverables from the structural design team.
And I am having four deliverables from the electrical and mechanical team which is;
foundation layout and sizes, access for flooring, access for finishing, handover to client.
So, there are four deliverables as a result from the electrical and mechanical team. From
the contractor there are five deliverables which is what we have come across handing
over to client, architectural drawings, access for first fix items, access for second fix
items, integrated testing and commissioning and sorry this handing over off to client is
coming from the other team as well.
So, now, let us see the BDM representation here. So, these are the same example I have
shown it, the same example I have used here, so that is easy for you to go ahead because
I am using only symbols here for representing because of the too many texts written on
the slide.
This is my structural design team, electrical and mechanical team and this is my
contactor team, so these are the three teams I have. And all the deliverables which are all
shared between the teams are all listed here and there is a short abbreviation used for all
these deliverables; the time duration or linkage between these deliverables are all shown
here.
Now, let us move on into analysis, because you are more worried about how to do my
scheduling with information exchanges, so let us see how to do the analysis.
Same way like your earlier network analysis; you will have forward pass, backward pass
and float ok. So, in the forward pass there is a formula written and again let me also say
something. So, in the BDM network analysis just like your PDM network analysis you
can still do two ways of analysis; one is the activities can be interrupted and the other
cases activities can be contiguous.
In the particular method where I have written all the formulas and I am going to explain
you with an example we are assuming: the activities are interruptible in the middle which
implies the activities can be broken down as and when the progress can happens on the
critical path and accordingly calculated. Equally in other method of contiguous execution
is also possible and the forward pass, backward pass calculations are accordingly
different. And you will have a different formula for the same.
So, I am not going to narrate the story on how these equations have come; if you are
interested you can go through to search papers and you can still see ok. So, we are only
going to see the equations, explanations and how do we apply in a project ok. So, this is
EET which is implies early event time. So, what we are going to do is, we are going to
take a simple example and at the point of the linkages.
So, wherever there is a link, we are going to break down this the highway design in two
different sub parts. And we are going to treat this as sub-activity of the highway design.
So, I am having highway design sub activity 1, sub activity 2, sub activity 3, 4, 5 and so
on. So, accordingly you may have to stop that.
And if you want to find the time at which these links are coming in, then I may have to
treat it in terms of nodes. So, I am going to work in terms of events only, the nodes
primarily to me are my events. So, I am going to work on event calculations.
So, what happens is EET j which implies early event time of a successor, which is equal
to maximum of the first EET i will be 0 or it is maximum of my earlier links only. And
the EET j is what you have to calculate, which will be maximum of EET i plus duration
of the sub activity or the maximum of the two values and the other one is EET of my PL.
So, what are these terminology? So, EET j is early event time of the current node which
is at the jth position not at the ith position. And the EET which is equal to early event
time, so primarily the start of the particular event plus duration of the sub activity in the
same team and or it is maximum of the other value; the other value is primary the early
event time of all the preceding nodes through the preceding links ok
I may have too many predecessor or linkages coming in at that point. So, you have to see
the early event time of all the preceding links. So, with that maximum of all these values
you will take that value as my EET j.
In the backward pass again, it is called late event time of i; because late event LET j is
primarily the maximum value you are getting with the, you are getting at the end for the
last portion you will take the maximum of your EET j. And for the previous activities or
events you will take the minimum of the values or which has come from the preceding
links.
So, accordingly what you will do you will do minimum of LET j, which is from the last
node you are traveling minus duration of the particular sub activity you have taken into
consideration or LET of my successor links. So, I may have too many succeeding nodes
coming out from the same late event, you may have to see that and you may have to see
whichever is minimum that you may have to take it for my a late event time.
So, which is minimum of LET j? So, the LET j is late event time or the succeeding node
or event along the same team minus duration of the sub team between the node i and j.
And the LET SL is late event time of all succeeding nodes through the successor links.
This may look complicated with an example I think you will understand much better.
The next is float; float calculation formula is very clearly given, late event time of the
successor link minus the early event time on the predecessor link minus duration of the
activity; following similar to your PDM analysis.
So, LET j is late event time of the end node or event of the current sub-team and EET i is
early event time of the start node or event of the current sub team.
Now, let us move on here I have taken a simple hypothetical example; all these formulas
I have written in one side for you to cross check and then go ahead with the calculations.
A’s duration is 4, C’s duration is 5. So, now, we are starting with the links; the first link
is coming from A to C. And the linkage value is given as 2 to 1, which implies A1, the
sub activity A1 takes 2 days of time and then it is releasing this and the sub activity C1 it
is taking 1 day of time to release my to receive my link from A.
Then there is a next link from here. So, this is actually a truncated portion. So, now, my
next links from here. So, at the time of 5-4 which implies; from the starting point, the
duration of this is 5 and from the starting point, the duration of this is 4. That is how the
notation reads. So, which implies, now what is the duration of my sub portions. So, C2 is
a sub activity here, duration of the sub activity is 4; A2 is a sub activity with the A and
the duration of the sub activity is now 2 ok.
Now, let us move on with calculating my forward pass and backward pass ok. Now let
me erase this, so that otherwise it will look, it will be overwritten. Now, so the further,
let me start with A and early event time I have given some notations here only for
understanding, only for me to note put it in a tabular form; otherwise there is no purpose
here, you can directly put the values here.
So, this EET for this particular node. For this particular node the EET value is 0; EET for
this particular node is given as, this is 0 plus duration of this activity is 2 ok, I am going
to use this formula here. So, this is 0 plus 2, any predecessor link is coming into this,
there is no predecessor link coming in here. So, this is 0 plus 2 only and that is why I
have written EET for this particular places 2 ok.
Now, I am going to start from activity C, because the links are coming from opposite
side. So, EET for this particular place is 0 and for this particular link my value is 0 plus
1. So, which is equal to 1 and I have another segment here EET of the predecessor link;
there is a link which is coming in to me and EET value for that link is 2 ok.
So, now what is written here maximum of these two values, so 1, 2 maximum value is 2.
So, my EET at this particular place I have, I am getting an EET value of 2 at this linkage.
Let us move on into the next link so; obviously, this is a successor for me and this is my
predecessor. So, I have to start working out from here. So, now, what about this value,
what about this early event time. So, early event time is, this value is 2, 2 plus 4 is 6
there is no predecessor. So, this value I am getting it as 6.
Now what about my early event time for this particular node. So, I am having 2 here, 2
plus 2 is 4; with the help of my predecessor link the value is 6, maximum of these 2
values is 6 for me ok. So, I am getting a value called 6. Now let us erase this, so that it is
easy to show the late event times.
Now let us see the backward pass. Backward pass we have stopped at this particular
point. So, this will be my starting node with which you have to start the backward pass.
So, late event time for this particular event will be 6. Now next will be this particular
event, because the link is traveling from this side. Now what will happen here is,
minimum of late event time minus duration, there is no successor activity here. So, you
cannot take this. So, this becomes nullified here and what about the late event time of the
successor link, I am having a value called 6; that is why you got a value called 6 here.
Now moving backwards, so this particular event. So, late event time is, either I should do
6 minus duration 4 which is equal to 2 or any successor links which I am having. There
are no successor links here, so this value will be noted down as 2. Now let me move this
way, because the link is travelling here.
Now what about this particular place, so I am at this particular place? So, the value is late
event time 6, 6 minus duration of this particular activity is 2 which is equal to 4; and the
late event time of a successor link ok, because I am having a 2 here, so 2. So, what I
should do, minimum of these two values 6, 6 minus 2 is 4 here and because of this link I
am having a late event time of 2. So, among these minimum value is 2. So, this value is
now taken up as 2.
Now I can do this or this, anything is fine ok. So, I am going to do this first; late event
time of this particular node is 2 minus 2 is 0, late event time of this particular node is 2
minus 1 is 1. So now, forward pass is done, backward pass is also done.
Now, let me erase. Let us see the float calculations now, float calculation is very simple.
Let us go through that.
Float calculation, so it is late event time of 2 minus early event time is 0, so minus
duration of this is 2. So, this value is now 2 ok. So, this is 2 minus 0 minus 2 which is
equal to 2 ok, for this particular segment.
Now let us take this particular segment, float of this particular portion. So, float of the
sub activity is what I have written, I am going to take the float of sub activities. So, late
event time is of the jth side is 6 minus the early event time of the same segment is 2
minus duration of this is 2. So, I will have 6 minus 2 minus 2, so that will be equal to 2
here.
Now what about this particular segment? So, this is my sub portion here, late event time
is 6 minus early event time is 2 minus duration of this activity sub activity is 4, so 6
minus 2 minus 4. So, this becomes 0 here. Now let us look at this particular portion. So,
the late event time is 2 here ok, the early given time is 0 minus duration of this particular
activity is 1. So, this is now will be 1 here ok, that is how the calculations are done.
Now, so far we have seen forward pass, backward pass and also we have seen the float
calculations also. Now we have finished up the entire network analysis ok. As per the
analysis now what happens? So, this particular activity is taking 0 float, this particular
sub activity is taking a float of 2, this particular sub activity is taking a 0 float, this
particular sub activity is taking a float of 1 ok.
So, now what are my critical paths and how do I show my critical path? My critical path
here is A1 with the link and C2 is critical path here; because of the 0 floats I have on this
particular network ok that is how you have to complete the analysis. So, network
schedules, how to represent the BDM representation itself you have understood, only for
one linkage we have seen; one type of linkage which is a N1 to N2.
And using the three formulas primarily on forward pass, backward pass and float, we can
also show the individual segments, where the path is critical.
So, you should not have the impression that, every time I should be taking up only the
teams and parameters. So, far in this example we have seen multiple information links
travelling between parties, it can be any link between any two issues. For example, first
example I have shown, parameters travelling with the teams and in this example
deliverables traveling with the teams; say it can be any link, the link can be anything just
like in types of DSM. Here also it can be between teams and components, it can be
between teams and parameters, teams and deliverables, or components and deliverables,
components and parameters. So, based on whatever you want you can accordingly
develop different DSM’s of your choice.
Now whenever I make this information relationships, you can this is more like a bar
chart. When we went to the experts and ask them for opinion and inputs on this BDM
example, they were very happy to receive this BDM network ok, under so many issues
behind. One is this BDM really looks like a bar chart, but with lot of information stored
in. So, it is not like a basic bar chart, but it has so much of information contained within
the bars ok. So, they were very happy to visualize the entire network, by travelling like
this ok. After this, this deliverable, then after this deliverable, so they were able to see
and visualize the entire network, more than the network schedules on CPM or PDM and
so on.
So, they overwhelmed and receive this network with enthusiasm ok; but there are some
problems in implementing in the industry, because of the manual method with which we
are still working out in calculations ok. Still now, there are no software which can help
us to do all the calculations automatically, as you can see I also have done mistakes
while showing the calculations.
So, same thing can happen because of the large values you have with the information
flows and sometimes it may be even in hours and some values may be in when you are
converting in days to hours you may get large values like this thousand and so on. It is
really difficult for somebody to go through with the calculations that you have to keep in
mind; that is the only one drawback we have in BDM.
So, far we have finished DSM and BDM. So, information management, information
driven schedules; two schedules we have seen, one primarily in iterations, this I can see
it is partial and interdependent only, it cannot support iterations ok. It can only support
interdependencies, but interdependence is multiple information exchanges in two ways.
So, two different technique which, one can support till iteration and one can support
interdependence in multiple links. So, I have shown this.
Thanks bye.
Lecture - 12
Other Scheduling Techniques
Welcome to all of you. So, we are towards the closing of this course, so the last, but one
lecture is Other Scheduling Techniques, ok.
So, now let us start with bar chart. So, bar chart now and then I have discussed here and
there, but we have not given a proper introduction on to what is this bar chart and what
are the other varieties as well. So, bar chart, the name has come because of the shape of
the bar, the shape of the activities are shown as a bar which is like a rectangular shape
and hence the name has come as bar. And the other name is Gantt Chart, because of the
person who introduces bar chart is Henry L Gantt, that is the other reason why we call
this as a bar chart.
It is a most frequently used tool in construction and because of the simplicity, ease of
preparation and easy understandable format, it is very widely used. Even today many
construction firms are very happy in showing a bar chart representation. Now let us take
this example, ok. So, here if you see, the X-axis primarily shows the timeline, so these
are all the duration in weeks, ok; and the y axis primarily this is duration in weeks, and
the y axis shows the list of activities. For example, site preparation, it starts from 0, it
takes to 3 weeks to finish ok; after the site preparation is over, footings starts and it take
another 4 weeks to finish, ok. So, like this the bar chart representation goes on, and then
foundation, temporary electric service can start in parallel. After this height preparation,
we have water and sewer tap. Then after foundation, we have soil treatment, after then
we have framing, so then masonry walls; so this is primarily a structure built, an
architectural structure built for the construction ok.
Now in this example, the drawback here is; it is not used as a proper document for in
case of disputes ok, only the CPM schedules or PDM are used or they are considered as
an efficient document in case of disputes and so on, Because no relationship between the
activities are shown here ok. And the other problem is sometimes you also when you are
updating on maybe some activities have slipped or extended and so on. It is very difficult
for somebody to update on the activities unless you know the predecessors and
successors relationships ok; because network as such will not help you to do that because
there is no relationship shown between the activities ok.
Now this x axis I have told, y axis also I have told, it is a list of activities; x axis shows a
timeline and the length of the bar represents a duration of all the activities ok, that is the
main significance of the bar chart. So, it is a visual or a graphic 2D representation, it has
only two axis x and y axis of a project activities. And the sequence on a time scale bar
with no links generally shown between the activities, there are other varieties which
shows; but the conventional bar chart will not show any relationship between the
activities ok.
Now, activities representation generally it is a continuous bar, that is what we have seen;
but bar charts can also show non-continuous activities what we call it as an activity splits
ok. If you see here footings, it actually starts on 7, then at the time of 4th to 5th week,
primarily in that week there is a gap ok; which implies no work on footings happens and
then the work resumes in the next week onwards ok. So, you can show with a blank
space, so which is like this. This notation is also possible; otherwise, I can also show like
this. This notation on bar chart is also possible and the other way to represent the activity
splits are the necking of bars, which implies I am going to make an activity like this, in
the no show period it is a very slender bar and then we have another rectangle piece ok.
That is the way we represent the activity splits in a network ok, in the bar chart
representation.
Then, next is Milestone chart, you know what is a milestone, we have introduced in the
first week itself. So, milestones are events which does not consume any duration nor any
resources. Milestones are significant events which just gives you an information, that
something critical has to be completed or started on a particular date and so on, so that is
an indication on milestone. And bar charts can also represent milestones and what we
called it as a milestone chart. So, this example you remember very well, this was a
development of a computational laboratory and there we talked about milestone planning
and so on ok.
So, primarily, when the agenda is to set up the laboratory for January 2020s teaching
session; then everything from the, all activities regarding the computer, setting up
laboratory has to be completed before December end that is the plans actually. This is
broken down into 4 quarters ok. Now, if you look at this. So, primarily if you go from
the last, so software procurement, software installation and then training has to be done,
then only you can start the lab sessions ok; training for the tutor, staff, the faculty
whoever you want on important software’s are required.
Now, software procurement, I have marked a milestone here. This talks about the class
oriented software procurement. When we discussed on this problem, we discussed there
were two things, two classifications; one is primarily for large number of users for class
teaching and other one is few users who use those software only for research purposes
ok. So, now, for the class teaching, so there were few software and this procurement has
to be finished somewhere in the first week of December. Installation has to be over in the
third week of December that is the indication I have given. Last week primarily we can
plan for training, so that the entire lab can start by functioning by Jan. So, there are some
milestones and these milestones can be used in the bar charts also.
Now, coming to next hierarchical bar chart. So, this example you are familiar with is a
hypothetical example. The representation of the same example and a bar chart I have
given here, but general representation on a bar chart is, the activities has to be
represented or arranged based on the order of early start. So, which implies the C has to
be coming here and then we have to have B D; but there is another purpose of putting
this bar chart, I primarily wanted to show the hierarchical, so I have kept it like this ok.
Here, I am just wanting to show that there are some hierarchical bars for activity A
which is like this and for activity D ok; so this D is like an aggregation of several
activities and A which is taking 6 days duration is an aggregation of several activities put
together, there may be several sub activities, relationships, different durations and so on.
So, bar charts can also show multiple hierarchical next level layers of representations as
well ok.
Now next is Fenced bar chart. The main drawback in bar chart we have already told, they
fail in showing up the relationship between the activities ok.
So, now this is a Fenced bar chart. So, fenced bar chart just use a straight lines, so as I
have shown here in red color. So, from A to B there is a red line ok, which is seen here
and from for an activity D I have again taken an A here and to show the C, because there
is a float value. So, there is a dash and then I am again using a C here. So, that is
primarily visible here ok.
Now, this is primarily the fenced bar chart. Now let us talk about the Linked bar chart. In
Linked bar charts what happens is you primarily use arrow marks to show the
relationship. So, in a way it is easy to understand rather than the fenced cases ok. Linked
bar charts we will see there is an arrow mark from A to B, this shows A and B are
dependent and then I have a straight line here and then again from A to D and from C to
D, I still have.
Now, here you may have a confusion; the arrow mark from here, is it from A or from B
or from both that you do not know. So, there is a procedure called cascading of networks
ok. So, primarily what happens is you can rearrange the starts of activities in such a way
that it just coincides and forms a shape like a cascade form.
So, you can flow the activities like this from both the ends and then it starts merging up
just like I have shown here. Now A gives information to B and there is one more arrow,
so this A also gives information to D and C is to D is shown here, ok. This is primarily
called linked bar charts.
Now, primarily in your MS project, there is a bar chart view. MS project always
represents CPM or PDM network, this is a slab concreting work in hostel ok; but there is
a bar chart or Gantt chart view also. I think people who worked in scheduling techniques
know well about this. You can also do resource leveling, resource loading, you can also
show cost control on projects, everything you can do and there is a Gantt view on the
whole MS project which is like this.
Now, this Gantt view if you see, it is more like it is more or less looks like a linked bar
chart only ok. So, for example, so this is primarily the aggregated activity or we call it as
a hammock activity. So, the part 2 of concreting, then I have part 3 of concreting ok. So,
in the part 2 of concreting, column reinforcement, shuttering, concreting, deshuttering,
these are all the resources I need for all these activities. This is only to show the bar chart
representations and the software also I am introducing now then and there in parallel ok.
So, MS project predominantly we use is very eligible, it is good for CPM, PDM and or
your bar chart views ok. Now let us move on to Time scale diagram. So, I have a small
example, time scale diagram is very simple, the representations of a network schedule.
So, primarily it is a CPM schedule, representing a CPM schedule on and timescale is
primarily what we call it as a time scale network diagram. In a way we call it as AoA
representation on a time view ok; because, only on activity on arrows, you represent the
activities on the arrows, so it is just combining like bar chart and AoA views on to the
time scales ok.
So, this is a bar chart which you; obviously, know ok. Sorry here also mistake C has to
come here ok. Now this is AoA representation. So, I have activity A and C are in parallel
ok, it is they starts in parallel, this is activity AoA representation and after activity A, B
starts and after A and C, D starts; with the help of a dummy you are showing that logic
representations. Now what happens in time scale diagram, this is a time scale diagram,
the same relationship I have shown 0, 5 and this is 10 and this is exactly 11 here ok.
Now, if you want you can arrange all the critical activities on the bottom and non-critical
activities on the top or you can rearrange it in your own form ok. Now activity A is
shown here ok. So, which is like this; this implies A takes from 0 till, till this is still 6 ok.
Now if you see here C, C is starting till here, it is less than 5 still here, so this takes a 4
day duration and B takes a 2 day duration which is shown here and after A and C is also
shown with the dotted line. So, I am having here two links are coming in, so this is
activity D and this is an activity D relationship and it is shown on a time scale view like
this ok. So, this is all about on a time scale diagram and purpose of using, so only few
people use this time scale diagrams because bar chart is very convenient for people with
the linked bar charts ok.
Now, the next technique, we are just moving on. So, if you look at the history also, let us
see first bar charts came in existence and then your network schedules on AoA first
came, then AoN s, after then little later only PERT, PDM all came in existence ok
PERT is the expansion for PERT is Program Evaluation and Review Technique. PERT
was developed mainly to assist in uncertainties in duration. In CPM schedules if you see,
the duration is always deterministic. What do I mean by that is, the activity duration
which is consumed by any activities is always given in single number, deterministic
number and there is no uncertainty in the duration on CPM networks and all.
In PERT what happens is, it allows uncertainty. So, primarily the uncertainty is covered
in three forms ok. Three time estimates are given; one is optimistic time, then most likely
and the pessimistic time. And because of this uncertainty, this PERT also used for doing
probabilistic estimation on critical paths, sub critical paths or project completion time; it
can also help you to judge or estimate on the probability completion on any activity and
so on.
So, now, let us see optimistic time. So, it is a shortest possible time for completing an
activity, if everything proceeds as planned without any problem ok. So, if everything
proceeds as planned without any problem and here the activity is performed under ideal
conditions ok; no breakdowns, no delays, no waiting for anything, engineers come,
labors come, everything starting in time. So, such a short time for completing an activity,
what we call it as an optimistic time ok.
Most likely time, it is a time for completing an activity under normal conditions. So, in
this case the conditions are not ideal, minor mishaps may happen here and there; for
example, labors always come 10 minutes late that happens on almost on all the days of
the execution. So, like this all normal scenarios generally happen ok, waiting time, this
that and the time duration for all those waiting on it happens in the same way.
The next is pessimistic time, it is a maximum time required to complete an activity under
abnormal or extremely adverse conditions in which everything goes wrong. Let us
assume, the labors were not even available, they could not even come to the site, it was
heavy rain, this that. So, something happened only on the particular day and it got too
much of delayed. So, it is primarily the maximum time which has taken. It will not have
occurred earlier days in the execution ok But here this estimate does not include natural
catastrophes or artificial catastrophic like fire and natural issues like earthquakes, floods
etc it is not that pessimistic nature ok.
So, what happens when you are having a duration estimate in three times; primarily the
optimistic, most likely and pessimistic, what do you do, the activity duration in PERT
you generally convert into your one time value and then you still go on in executing the
calculations just like how you did in CPM ok. So, the activity duration in PERT is given
by te equal to plus 4 tm plus tp by 6.
So, the most likely time it takes the maximum number of occurrences ok. So, you are
taking 3, you are taking 6 times I estimate; one time on optimistic, one time on
pessimistic because it rarely happens and 4 times on the most likely and divided by 6,
there is a statistic background for all this. Anyway I am not covering all those, but only
giving you the formula now, and I am not also going to talk about the project finish
probabilities and so on.
Now, this is a simple example I have taken, there are 6 activities in this example, so
excavation, construct foundation, erect building structures, set equipment, install MEP
and roofing and complete the finishes ok.
So, the three time estimates for all the activities are given here as to, tm and tp. And I
have also calculated the te ok. te, I have calculated with the formula to plus 4 tm plus tp by
6. So, which implies for an excavation 12 plus 4 into 14 plus 16 by 6, so I got this as 14;
7 plus 4 into 10 plus 13 by 6 this is 10; 15 plus 4 into 18 plus 21 by 6, this is 18 and the
IPAs for all these activities are also given ok. Now, if you look at textbooks earlier to
1990 s, 2000 even, you will find most predominantly PERT networks are always shown
on AoA ok; activity on arrow networks.
But after then researcher said you can use any form of the notations either AoA or AoN.
So, I have used to shown it in AoN network ok. So, as I told you the 3 time estimates
primarily you are going to convert into one time estimate the t e values, expected time for
completing an activity. And if you are working on AoA you will still calculate early
event times, late event times, slack and then you will find the duration. If you are doing it
on AoN, then you will do all the float calculations just like your early start, early finish,
late start, late finish of activities; and total float, free float just like how you did for AoN
analysis.
So, just to recollect your memory, so, I am going to work it out. So, this is a AoN
network. So, A is the starting activity. After A I have to do B and after B I have to do my
C and my D, then after C and D is my activity E, then my activity F, durations are all
dumped ok, C is 18, D is 29.5, this is 9, 7, I have written. So, now, this is a forward pass
I am going to do, so this is 0, 0 plus 14 is 14, so this is 14; 14 plus 10 is 24 here, so this is
now 24; this is also 24 ok. So, now, if you add, so this becomes 0.5 13 1, 53.5 ok.
Now, here it is 12 1 42 ok, now MAC back. So, E is early start of activity E is maximum
of early finish of these two activities, so this is 53.5; 53.5 plus 9 becomes 62.5. So, now,
this is 62.5 and then this is 69.5. So, now, working backward pass which is the late starts
and late finishes so, this is 69.5 ok, 69.5 minus 7 is 62.5, so this is 62.5; 62.5 minus 9 is
53.5, so this is 53.5 and this is also 53.5; 53.5 minus 18 is, so this is 0.5 this is 5, 4 minus
1 is 3 and what about here, so this become 0. So, 13 minus 9 is 4 this is 24 ok.
Now the minimum of these two values is 24, so I have to go with 24; 24 minus 10 is 14,
so this is also 14; 14 minus 14 is 0 ok. Now if I want to do the total float calculations, I
am going to change my pen color now, ok. So, 0 minus 0 is 0 ok; 14 minus 14 is 0; 24
minus 24 is 0 and 53.5 minus 53.5 is 0; 62.5 minus 62.5 is 0. Now this is 0.5, 1, 11.5, is
my total float. Then the critical path is activity A, then I have activity B, then I have
activity D, and then I have activity E, and then I have activity F.
So, the same way you have. So, primarily the pert you are converting into deterministic
times on expected; so te is expected time of completing an activity and then you have to
go with the same procedure like how you did for CPM calculations ok. So now, we have
discussed so far on non-network schedules bar chart comes in that category, the various I
would say the other updates or improvements in bar chart also we have seen. The next
we saw PERT, so primarily it is close to CPM. So, CPM and PERT we have seen what is
the difference. So, CPM, pert, PDM also we have discussed in detail.
Now, let us move on into advancements in these techniques ok. So, CPM, the other one
is called GPM ok.
GPM the abbreviation is graphical path method ok. These techniques I may not be
showing you duration calculations and scheduling and so on, they are not that powerful
techniques; but I am just giving you an awareness that there are so many techniques
available in scheduling and you should know what are all available. So, GPM it is a
network technique again that offers the simplest possible schedule in the shortest
possible time ok. So, the abbreviation is graphical path method.
Now what happens here, so this GPM it has come after your primarily on the CPM
network only; and what happens is when you are showing your calculations on total float
and you are calculating on a critical path and so on. And once the project is progressed
and during the as-execution period ok, half of the project has progressed and so on. And
if you want to know what has happened in the as-built portion, then it is very difficult for
somebody to look at the updates and do analysis on the as-built portion ok.
The GPM is a method which helps you to analyze the as-built portion and helps you to
identify the float places, drifts and so on where are the buffers available, what has
happened, so it helps you to analyze where any mistake has gone and so on ok. That is
the primary advantage of using a GPM; the ability to determine total floats and the
critical path for the as-build portion of a schedule primarily the left of the data date, ok.
When you are working on project controls you will know what is a data date; data date
other name is called status date, it is a date on which you take updates on the project
progress ok, that is primarily determined by the CPM.
Now CPM is widely used; now these are the steps I would say on the building of the
GPM. So, CPM you know is widely used as a project management tool for planning and
scheduling. And the basics to CPM is primarily you have to draw a network and from
where you calculate the project completion time, total floats, critical path, etc. primarily
you do network schedules and also you do network analysis on AoN network ok.
Now once the actual dates are introduced in the CPM network like maybe, you plan 6
days it takes, but what happens is the first date may be on 1st of July and the sixth date
may be on 7th of July when the actual dates are implemented and the as-built drawing is
really generated. We start losing out the information on the data date, which implies a
previous data date like total floats and where are all what happened the flaws, where are
the delays happened and so on. We generally, we do not take a notice of, we miss all
those and the critical path can no longer be calculated. So, in for those instances, GPM is
very widely used ok.
Now, if you see there are certain terminologies we have specifically for GPM; number
one is called Drift this is not the complete list there are so many terminologies like this.
Number one is drift, the amount of shift to earlier planned dates while not forcing an
earlier project start it is primarily called as a drift. The next is Float, the amount of shift
to later planned dates without causing an overrun in a project completion. So, in a way it
is like a total float sort of set up in your CPM, we call in GPM, the name given in GPM
is float. The next is Drift Buffer the amount of shift to earlier planned dates without
forcing any of it’s predecessors to earlier dates ok.
So, primarily what happens is, there is an end point software for operating on GPM. So,
when I am shifting an activity it tells me how much I am shifting and because of my
shifting pre-data date ok; is there any predecessor activities or should be deviated or it
gives me a caution as to the predecessor activities have to be also preponed and so on ok.
So, accordingly you have to see how much is a drift, how much was a buffer these
activities have consumed, we can still explore.
Next is buffer the amount of shift from planned dates without delaying any of it’s
successors from their respective planned dates; in a way it’s like your free float we have
in CPM terminologies; then we also have other terminologies like offset, embed, forensic
floats and so on ok.
Now, let us move on, because my purpose is only to introduce these techniques to you
the next is GERT. So, GERT similarly like GPM it is an advancement to PERT ok. In
PERT you know, why we introduced PERT; PERT has come into existence only to allow
uncertainties in durations ok. So, GERT has developed by Pritsker, it has developed as
means to advancements in PERT ok. So, in GERT also we can allow deterministic
duration as well as probabilistic duration estimates just like your PERT. So, GERT is
same as PERT in that sense; but GERT is different from PERT in two ways, one is it
allows loops ok. If you take any of the network schedules like your CPM, PDM, PERT,
etc, they does not allow loop even your GPM; but GERT can allow loop because of the
probabilistic branching and so on.
And GERT also allows probabilistic treatment of network logic and branches. In the
sense it can also have two links and one of the link can be operational and the other one
will operate only at the certain situations or circumstances. So, it can allow either one
link to operate, both the links to operate, so it allows branching off that is what we call it,
as a branching of a network, with example you will understand later. Then GERT
networks include directed branches ok, which implies arcs, edges and transmittances and
logical nodes also which we call vertices. So, two parameters are linked with the branch;
one is a probability time and the other one is the time taken, time required for the branch
to take.
Now, there are certain terminologies in GERT ok; one is called input terminology and
the other one is output terminology. And how do you draw the network, because I told
you the network itself is very probabilistic and so, the network itself should be you know
guided as to, whether this arrow should work out or this arrow should work out. In CPM
if you draw two arrows, both the arrows will be operational at every point of time ok; in
GERT it need not happen. So, logical relationships on the input, now let us see the
symbols and then we will see an example.
So, when I am drawing a straight line with a triangle it is called Exclusive OR and one
and only one of the branches leading into this node can be realized at a given time. So,
only one branch leading from the node will be realized here ok. The next is Inclusive
OR, the realization; so primarily it is only a triangle shape, the realization of any branch
leading into the node causes the node to be realized, ok. So, whatever comes in it just
does it’s operations ok.
Next is AND, the node will be realized only if all the branches leading into the node are
realized. I may have two three links entering into this set up. So, what happens is, all
when all the links are having a ready for start then only this branch will work; otherwise
it would not work, so this is the input.
Now let us see the output symbol. So, primarily I am having two symbols here ok; one is
like a D shape and the other one is a triangle shape. So, D shape stands for Deterministic
and the triangle stands for Probabilistic ok.
Now here the D, all the branches emanating from the node are taken if the node is
realized ok, so all branches will be operational here. Then Probabilistic, exactly one
branch emanating from the node is taken if the node is realized; whenever there is an
enough input coming in only one branch will be operational ok. Last three cases we have
seen only on the input, so I am talking about arrows coming into the network. In this case
on the output when you have this symbol in the end, I am talking about arrows starting
from the particular network.
Accordingly, now I have several combinations here, you will see there are so many
combinations here, ok. So, this stands for input and this stands for output, in all the cases
this stands for input and this stands for output, ok. If you see here, so you cannot use this
separately, you have to use combinations only because you need to know what is the
input criteria for taking up the network and what is the output criteria ok. If you see here,
the first one is primarily exclusive OR and deterministic; the next one is exclusive OR
and probabilistic and the third one is exclusive OR and deterministic; then the next one is
inclusive OR and probabilistic; then next is AND and deterministic AND and
probabilistic ok.
Now, let us take an example, I would say this is a gas station construction which happens
in a hill station ok. I would say that landslides are more in number, this network was
drawn with DSM; but primarily what I wanted to show here is there are three
uncertainties which are encountered in the entire network.
Whenever you have an uncertainty only then you want to draw a GERT network. So, we
have taken these three as uncertainties here; number 1 is on PCC plain cement concrete,
number 2 on bar bending and reinforcement, number 3 on concreting and compaction.
So, it is primarily excavation of breaking of rocks and then disposal of rocks, then PCC
primarily pouring of concrete, then bar bending or reinforcement, cover blocks, then
shuttering, batching, concreting compaction and curing ok.
What happens with the hill station with a lot of landslides; every time you may have to
redo the work again if the landslides have occurred, maybe you have done excavation,
landslides have happened, then you may have to RE-clear the site for the next level of
excavation, ok.
So, this is the simple network which we have taken it up and the representation on GERT
looks like this ok. Now these are the symbols which I have used earlier in combination
symbols; every time you may have to use this combination symbols ok. Only in the start
or end you may use a single symbol ok. Now let me write everything, so that it is easy
for us to understand. So, this is exclusive OR and deterministic ok, this is exclusive OR
probabilistic this is inclusive OR deterministic; so this is inclusive OR probabilistic; this
is AND deterministic this is AND probabilistic ok.
So, I am starting and what shape is this, this is AND probabilistic, this has come because
of the link I have in here. So, now, let us start with the activity, do not see the symbol,
the symbol will be operational only when the back link works ok. So, I have two arrows
coming in; one is excavation and site clearing and the other one is rock breaking and site
clearing ok.
So, I have two modes, either I may have to clear up my earth by excavation or I may
have to do rock clearing also. So, that is an uncertainty coming in ok. So, AND implies
all the links coming earlier has to be considered and probabilistic is, so either one of
them will be operational ok; either I may have to do excavation or I may have to do the
rock breaking and unsite clearing. Because I have told probabilistic end, I may have only
one link operational.
So, suppose I am taking excavation ok, there is no rock I am only doing excavation there
ok. So, again what happens as and when this link works this will start again I may have
either one of the links only; either I may have to do no PCC or I may have to do PCC
work. So now, what happens here is, there may be a problem with the pouring of
concrete itself because I told uncertainty happens at that place itself. So, I may have to
do PCC or I may also avoid my PCC ok; because of what there may be a landslide which
has happened. So, when the landslide happens, I may not be able to do PCC I may have
to reclear due to landslides and again I may have to repeat the process ok.
If the PCC happens in the sense, if there is no landslide, no uncertainty has happened,
then I am going to go to the second step in my operations which is the plain cement
concrete work. Same thing, instead of excavation I had rocks available, then I may have
to do rock breaking; and when the rock breaking happens, again there may be a landslide
which may happen. So, once either I may do no PCC or I may have to do PCC and then
do the land clearing process that is what.
So, if there is a land clearing, then I may have to come here and do the land clearing and
again repeat the steps accordingly. If there is no land clearing, if there is no landslides
then I may do with the PCC work ok. So, whether I have taken excavation work or my
rock breaking work, then I could do my PCC if there is no uncertainty that has happened.
Now there is again an inclusive or probabilistic link given here, ok.
Now either reinforcement work can happen or reinforcement failure can happen ok.
Maybe because of my quality of reinforcement or something, reinforcement failures are
more common their or maybe I have tried to put in the reinforcement in the excavated
portions and landslide has collapsed everything. Some uncertainties can happen with the
reinforcement also; in the earliest slide itself, I have shown you three places with the
study on previous projects we have identified that uncertainties can happen on these
three instances.
So, reinforcement work can happen or as I told you only one link may work or
reinforcement failures can happen. Now when the reinforcement failures happen then
rework has to be happen. So, again the reinforcement work has to be passed on or
because of the poor quality of reinforcements or if the landslide has happened and
collapsed my reinforcement, then re-clearing of the landslide has to happen, again
excavation and all these work has to happen and proceed. Then once the reinforcement is
done then we go on to shuttering work this is primarily the AND deterministic, for
shuttering I need the concrete line re-clearing and also the repair of the mixture or setting
of mixture. So, I need all the things to be ready and then only I have to go with my
laying of concrete lines ok.
Now, with the concrete lines I may again have a failure, concrete may happen when there
is rain or a snow, then I know what precautions to take; sometimes concreting may not
also happen with the hindrances ok. So, concreting happens and concreting has not
happened. So, this is concreting without hindrance and concreting has failed because of
the rain or snow ok. Now what happens here is, now this has gone to the target
accomplished stage ok.
Now if there is a hindrance due to local people or something then target not
accomplished and then this network has to go in loops. When you give the probability
value as to what is the probability that PCC may fail or your reinforcements may fail
according to the probabilistic values you dump in onto the network then what happens is
accordingly the durations can be calculated. But still no far has worked on duration
calculations on this, but it can help you to draw the network on probabilistic branches
and roots, so this was drawn up by Pritzker ok.
Now, moving on into, so primarily if you have seen, we have covered CPM and PDM
very much in the first few classes. Now previous to PDM or CPM we had the bar charts
that we have discussed right now, then we had PERT which we missed, advancements to
CPM also we have seen GPM, advancements to PERT is the GERT that also we have
seen. The next is your PDM relationships and advancements to PDM we can say it is the
BDM or your RDM ok, relationship diagramming method and beeline diagramming
method are generally called as an advancements, Beelines are to represent multiple
relationships. And when you want to have a resource dependency shown on the network,
then RDM can play a picture these are all called advancements.
Now in the BDM network itself if you want to represent more information, then some
researchers have developed their network called continuous and point to point
relationship ok, this was developed by Hajdu, Miklos Hajdu and what does he say is, I
am having an activity this is a legend with which you can understand, the legend has
name of the activity and duration.
So, this says activity A duration is 5, activity B duration is 10, activity D duration is 20,
activity F duration is 20 and there are different types of relationships given; point to
point relation it says primarily the 1st point, 2nd point and the time gap. In your BDM
you always had N1 and N2 and if you want to have N1, N2 with the time gap, then you
can mark this as F, S and O primarily the finish the first segment, the second segment
and then the time lead or lag ok. So, (F, S, O), (F, S, O) here there was no finish to start
and there was no time lag here no time lag here. So, after this activities done only you
have to start, but here after one day of activity you can give information to F and there is
no weight, so it primarily the start and then I have a 0, so there is no lag here ok.
Now here there are also written in terms of duration and distance ok. So, the X m it talks
about work gap and the X day it talks about time gaps. So, now when you have F S 0 on
the arrows it means it is a point to point relationship and when you are having an X days
or X m it talks about time gap or the work gap, which implies continuously between two
activities I am having a relationships here. When you are talking about repetitive
activities in for example, what happens is you may have let us assume I am doing
excavation with one stretch second unit, third unit and so on.
And suppose if you want to maintain a continuous interval between all these segments,
then I may have to maintain a time gap or it may be a distance gap ok. One meter ahead I
should not keep all my equipment and there should be a minimum clearance of one
meter, then you can write one meter there it shows a work gap. Otherwise one day later
the activities should progress, then throughout excavation and the second activity then
you can maintain the X day as a distance buffer ok.
So, this technique is very helpful in showing the repetitive activities also because what
these researchers say is, in reality you may have activities with continuous and non-
continuous put together. So, you need a representation which shows a non-continuous
segment also along with the continuous or repetitive portion of activities ok, for that case
this is very useful ok.
Next is LSM, LSM is linear scheduling method. In the regular class I explained you on
line of balance method, line of balance you have to draw 2 bars in order to show the
early primarily. The start of the segment and finish on the segment there is nothing called
early and late ok, which shows start of unit one finish off unit one, so it shows as a
rectangular bar and the entire progress till the last unit is seen here ok. In LSM what
happens is, it shows, so primarily this is a network it shows only one horizontal line and
with an inclination to the horizontal. So, primarily I have taken 3 activities A B and C
and this is a duration of the activities and this shows units 1 2 3 4 and so on ok.
So, I am having an activity, primarily I am having a rectangle here, this shows the time
taken with which the first unit is done ok. So, the 1st unit when I am drawing a line it
says I am finishing it at 3 days, the 2nd unit I am finishing it at 6 days, 3rd unit I am
finishing it at 9 days and so on. So, 0 to 3 ok, 3 to 6 that is how you have to represent or
interpret your LSM schedules ok. Now for here, there is a distance buffer also in the
earlier LoB also we showed the buffer by means of a start buffer or a end buffer, same
case here also we have to show the buffers in LSM .
Now what happens is, there are software which are very specific only for repetitive
activities. We are just moving into that.
But there can be non-repetitive portions also in a repetitive stretch of an activity

completions. This is not a new technique what we call as a time-chainage diagram; this is
a very old technique which came in as soon as your LOB was formulated ok. There are
so many terminologies came along with this then only the CPM form of a network came
in existence. So, these all techniques have come even earlier to CPM schedules ok. So,
time-chainage diagram it just talks about time plus chainage, now we are into repetitive
activities. So, I have duration as my stretch and I should also know where these activities
are going to be done in the whole chainage ok.
I am having a 200 days of my duration and when you are saying excavation it takes for
throughout stretch, but you should know on day 60 to day 80. This excavation is
happening on which stretch and where is the other activities happening at the particular
point of time, so primarily the chainage is the whole segment or distance, I would say
along with you are representing your time. It is widely used for all repetitive projects like
your road projects, bridges, tunnels and so on and it is easy to know the order of
activities, locational happening of activities and etc ok.
Broadly the researcher has classified four types of activity; one is a compact. Compact is
generally shown with the incline line ok. These examples I have taken with the help of
earthwork or, maybe kerb laying or fencing ok. Extended lines, I may have parallel
inclined lines here; this is primarily to show concrete paving. Maybe there is a duration
there and it is like a completely happening an activity. Then extensive activities are
shown as blocks for example, earthwork ok.
Earthwork can be done throughout the stretch of your highway construction ok, which
you may have to show it as an complete block till the chainage is completed. Static
primarily show it in vertical columns. In the entire highway segment, I may have bridge
maybe in one or two places only; bridges, junctions or tunnel shafts. So, only in few
places I may have the vertical columns coming in and to show the bridge construction,
peers, abutments and so on. And with that, I am closing my diagrams ok.
So, these are the main features in non-continuous representations. And there is a specific
software which helps to show your repetitive projects ok, what we call it as a TILOS ok.
So, TILOS is primarily the time linkage software and here you see, this is the same
example what we have discussed the earlier on highway widening project. So, this is a
snapshot on TILOS. So, I am having a March ok, April, then May, these are all the
weeks ok; I am starting on 10th week is somewhere in March, 11th, 12th, 13th, 14th,
15th and so on. So, these are the weeks and along with the year of representations.
Now the excavation work is starting somewhere in the 12th week of the project. So,
which is happening in March and it is finishing in somewhere in close to the 14th week
of the project progress, which is supposed to be the end of March towards the end of
March. This is the project progress week and this is your month, the relevant month on
which this entire work is going on. Then excavation, embankment, these are all parallel
inclined lines which shows your regular activities ok. If you see here, compact activities
are shown generally by inclines lines, right. So, these are all your compact activities.
So, first excavation, then exists embankment, subgrade, granular sub base, wet macadam
mix, then I have my DBM, then this is my bituminous concrete mix ok. So, all these are
shown. The crew connectivity is also shown for the next levels and so on, because I am
sharing so many resources between the activities, which during the RDM modes we have
discussed this also.
Now, let us move on in to see the time-chainage representations on TILOS. This is again
a TILOS snapshot only on the same project, where in I have raised a bridge also to
continue with my highway construction ok. So, these are all girders here; the girders are
primarily the extensive activities which are shown like blocks and they occupy the
complete chainage ok, almost most of the chainage ok.
Suppose if the entire construction is happening on bridge and highways, then you may
have to show all these on girders on extensive activities ok. These piles are all static
portions which happens only from this particular segment to this particular segment.
Here you will have a chainage also and I have to crop it because I did not have and I
could not put it fully.
Ok, below this at the end I may have the chainage also, which shows 0 kilometers, 2
kilometers, 2.7 kilometers, this may be 3.2 kilometers; all these can happen and you will
know so this, from this particular distance to this particular distance I am doing my
foundation. Foundation is happening from this time to this time, the second foundation is
happening from this time to this time and it is happening in the month of April 3rd to 4th
week. So, all these you can interpret on seeing the figure only ok. So, it is still a
visualization diagram and also an analysis tool on repetitive construction projects. That is
all about TILOS ok.
So, now, we are just wrapping up from the beginning. So, first non-network schedules
we have seen, network schedule CPM and PDM we have seen earlier. Now we have
discussed PERT, GPM, GERT and so on, and then we moved on into repetitive
construction projects. LOB we have discussed enough in detail in the regular class. Now
I have introduced to you LSM and time chainage charts and this TILOS can help you to
know to represent these networks also ok.
MS project or Primavera are very relevant for all your earlier schedules like your CPM,
PDM, bar charts and so on. And then, TILOS is very good for LSM. It does it will not
show LOB scales. It only use as LSM and time-chainage charts ok. So, in way repetitive
projects can be shown on this TILOS.
The next is types of activities we are just moving on into types of activities, to show
where we are and what all we have discussed so far. So, independent activity, dependent
activity we seen, we know what is it. So, your CPM, bar charts whatever you have
discussed so far will surely take care of independent and dependent activities.
The next is semi-independent; semi-independent is, you need not wait till the end of a
finish to start conversion relationship. If you are ready to send any pass on any
information or you are ready to start your work anytime in the midway of your previous
predecessor. Then you can use a relationship called semi independence activities ok.
Your PDM can really represent the semi-independent activities. This you can use CPM,
you can use PERT, or you can use bar charts. So, all these are helpful for that, I am not
talking about repetitive activities here ok.
Next is interdependent activity. DSM is a powerful tool for representing interdependent

activities we have seen that in very much detail. The next is multiple two way
information linkages or beelines and BDM is a very powerful method for introducing
and if you want to mix and use repetitions and BDM and so on. Then so, BDM is
primarily like point to point and if you want to have continuous relationships on distance,
time and so on, then you may have to use what they are, what professor Hajdu has used
on continuous and point to point relationships.
So, all these, you know what tools we can use and how to do the scheduling with all
these methods ok. But one thing you should understand these are all primarily comes
under the classification called natural overlaps; wherein, the information can pass on
easily at that time and you are scheduling only with that knowledge ok.
Suppose if you want to know force and activities to release information ahead of
something or behind something and then you wanted to schedule a project, then you may
have to do forced overlap ok. Forced overlap, I am not going to cover in this particular
course ok and interested researchers you may have to go through research articles only
on forced overlap.
So, all these are our natural overlaps and we have seen all these techniques right now ok.
So now everything has been covered, some in detail, some atleast I gave you an idea. So
starting from bar charts, you have all the varieties like milestone chart, fenced bar chart,
linked bar chart, hierarchical bar chart and so on. Then time scaled diagrams, network
schedules CPM, PERT, PDM, Repetitive projects: LoB, LSM, Time-chainage diagrams.
Other techniques like GPM, GERT, BDM, RDM, DSM which are extension.
MSProject and Primevera you can you
TILOS
NetPoint for GPM only
Beeliner Contigous
With this, I am stopping for today’s class ok. With that, we will wrap up with the course.
Bye.

Scheduling

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Scheduling

Uploaded by

Copyright:

Available Formats

Scheduling Techniques in Projects

Dr. J. Uma Maheswari

(Refer Slide Time: 00:33)

(Refer Slide Time: 03:59)

So, if you look at project management in construction industry or what we call it as a

(Refer Slide Time: 12:26)

(Refer Slide Time: 14:55)

(Refer Slide Time: 16:14)

(Refer Slide Time: 30:03)

(Refer Slide Time: 32:04)

(Refer Slide Time: 33:37)

(Refer Slide Time: 34:38)

(Refer Slide Time: 36:29)

Then prototype models also because it was supposed to be a construction laboratory.

(Refer Slide Time: 00:42)

(Refer Slide Time: 08:50)

(Refer Slide Time: 09:09)

(Refer Slide Time: 10:46)

(Refer Slide Time: 13:11)

So which implies any activity should be performed in a normal speed, in a normal

(Refer Slide Time: 18:13)

(Refer Slide Time: 19:52)

(Refer Slide Time: 21:55)

(Refer Slide Time: 26:56)

(Refer Slide Time: 35:01)

(Refer Slide Time: 37:26)

The next is redundancies; redundancies are unnecessary relationships in a project. I will

Constraint is when an activity is subject to constraint such as approval of an owner or a

(Refer Slide Time: 42:08)

The next predecessor activity and immediately predecessor activity; I am having an

The next, so what is a network diagram or a schedule? So, network diagram or a

(Refer Slide Time: 00:52)

(Refer Slide Time: 03:20)

Now Activity on Arrow diagram.

(Refer Slide Time: 05:01)

So, these are the general issues with AoA diagram.

(Refer Slide Time: 07:19)

(Refer Slide Time: 08:09)

(Refer Slide Time: 09:03)

(Refer Slide Time: 12:02)

Other option is; so two options are already shown here.

(Refer Slide Time: 13:54)

(Refer Slide Time: 14:30)

(Refer Slide Time: 18:29)

Now what do you do after arrangement? So, first is AoA diagram.

(Refer Slide Time: 21:52)

(Refer Slide Time: 23:08)

(Refer Slide Time: 24:39)

(Refer Slide Time: 29:11)

(Refer Slide Time: 30:26)

(Refer Slide Time: 35:27)

(Refer Slide Time: 36:43)

(Refer Slide Time: 37:45)

(Refer Slide Time: 44:16)

(Refer Slide Time: 44:33)

(Refer Slide Time: 48:53)

(Refer Slide Time: 50:43)

So, let us assume I am having a project which is planned to be completed in 30 days as

Suppose if the project completion contract completion time is something around 40

(Refer Slide Time: 00:23)

(Refer Slide Time: 03:41)

(Refer Slide Time: 10:23)

(Refer Slide Time: 11:31)

So, I am having an activity A of duration 2, I am having an activity B of duration 10. So,