Ihab Saad – Precedence Diagramming Method PDM
AI: Summary ©
AI: Transcript ©
Music,
welcome to another class in construction management, 324,
construction planning and scheduling. And today we're going
to start discussing another scheduling technique, which is
activity on node, also known as precedence diagramming method, so
either known as AON activity on node or presence Diagramming
Method, PDM, last time we discussed ADM activity on arrow,
or AOA activity on arrow, or ADM Arrow Diagramming Method, and we
learned about the network flow from left to right. And we learned
about the not the nodes that represent the start event and the
finish event of the activity, and the line connecting these two
nodes representing the activity itself. And we learned about some
of the difficulties representing activities that start at the same
time finish at the same time, and having a common predecessor or a
common successor. And we had to resort to using the dummy
activities in this case. And another problem with ADM is that
it only allowed for one type of relationship linking the
activities, which is commonly known as finish to start, where
the successor has to start, only after the predecessor was
complete. While this is not always the logic connecting the
activities in the construction industry, we had to look for
another alternative, especially also with the dummy activities not
being the most intuitive thing to understand. We needed to look at a
better way of representing activities. And that basically was
the reason why activity on node or presence Diagramming Method
started.
So we're going to discuss what's the activity on node network.
We're going to discuss the network calculations for AON or PDM
networks. We're going to look at the presence diagrams, and we're
going to look the precedence diagram calculations, which are
going to be pretty much the same as the ones in Arrow Diagramming
Method.
So this PDM, or presence Diagramming Method, also known as
activity on node network, instead of the activity being represented
by two nodes with an arrow connecting them, so all of these
three entities represented the activity the activity, we're going
to consolidate all the information about the activity into one box,
and that box is going to be the activity itself, rather than the
line. The lines, in this case, that you can see on the screen,
represent the relationship between the activities. So it uses the
node or box to represent the activity, as opposed to the arrow
used with the activity. On arrow or AOA network, the arrows
represent logical relationships, and their length does not reflect
any special meaning. So again, this is not drawn to scale. The
length of all the arrows are pretty much the same, so it does
not represent any special significance. And the good
advantage that we have in presence diagramming method is that we do
not have any dummy activities. So when you look, for example, at a
Gantt chart or a bar chart that has five bars representing five
activities, the translation is going to be one to one. Each bar
is going to be translated into one activity. We could not do that in
Arrow Diagramming Method, because, in case you had some activities
overlapping, or you had some lag between the activities, you had to
introduce another activity to represent that lag, or you had to
break the activity into more than two sections to represent the
overlap. So here in presence diagramming method, it's much
easier that every bar is represented by a box in this case.
So what we have here, for example, is a box representing the
activity, showing the ID of the activity, number 10, the name of
the activity mobilized, and the duration of the activity, followed
by another successor, or actually two successors, number 20 and
number 30 begin construction and side fencing two days and three
days respectively. And here it shows the relationship between the
activity. Looking at the box, we're going to find out that the
box has basically two vertical lines, the left side and the right
side. Since we already agreed that the network is always going to
flow from left to right. We're going to call this left side the
start side of the activity, whereas the right side is going to
be the finish side of the activity. Now looking at the line
connecting two consecutive activities, like 10 and 20 or 10
and 30, we're going to find that it starts from the right side of
the activity, which is called the finish side and ends at the left
side of its successor, which is called the start side. Therefore
this relationship is going to be called Fs, or finish to start,
because it starts from the finish side and ends at the start side of
the successor. And here we have the number zero.
Two which shows that there's no lag between these two activities,
which means activity 20 is going to start as soon as activity 10 is
complete. Now whether we add this zero or not is basically
redundant. So if we had zero, then it means it starts immediately. If
we do not add anything, it should be understood as exactly the same
thing. So the relationship between 10 and 30 is exactly the same as
between 10 and 20. It's still finished to start with zero lag.
We did not have to add that because, again, this should be
understood from the flow of the network. If, however, we add any
positive number or negative number. Here we're going to see
what's the significance of these numbers in a couple of slides.
The network should start with one node and end with one node, as we
have done before.
As usual, it flows from left to right, therefore the arrowhead is
redundant. Activities can only be linked either from the start side
or from from the finish side, or sometimes from both. We're going
to see that in a few minutes, we can never connect the activities
from the middle of the box. That would be a big mistake to connect
the activities from anywhere other than the start side or the finish
side. Of
the concept of legs and leads. Legs and leads are pretty much the
same thing, by the way, legs and leads if you're looking, for
example, at two
cars in a race car, and one of them is ahead of the other by,
let's say, a couple of yards, or whatever, in a very fast NASCAR
racing for example, you can say that the car number one is leading
car number two by two yards. Or you can say that car number two is
lagging by two yards behind car number one. So basically, the lead
and the lag are exactly the same thing, depending on it only varies
depending on where are you looking from. If you're looking from the
one that's ahead, you say we are leaving. Or if you're looking from
the one you're that's behind, you say we are lagging. So a lag is a
minimum waiting period between a start or an end finish
of an activity and the start or end of its of its successor. So
for example, a large concrete slab or the rebar can start after the
start of four work, but not necessarily wait for its
completion. I have a large concrete slab for a warehouse, for
example, a slab on grade, and the activities are going to be
basically four work for the sides of that slab, or even if it's a
suspended slab, not necessarily a slab on grade, if it's a suspended
slab, then we have the formwork, and then we place the rebar, we
place the mechanical electrical inserts, and then we place the
concrete. Now if it's a large slab, you don't have to wait for
the whole formwork to be done to start working on either the rebar
or the mechanical electrical inserts. You can start a little
bit after the start of the formwork, when you have enough
work to do. So you don't have to wait until the formwork is
complete. Therefore, in this case, we say that the rebar is going to
start after the start of the formwork, not necessarily after
the completion of the formwork. It does not have to wait until the
completion of the formwork. So in this case, the relationship is
going to be a start to start, but there's going to be some lag. You
cannot start on the same day. You have to not to have enough buffer.
You have to have enough work done on the formwork in order to be
able to start the rebar. So the arrow networks cannot accommodate
a lag, and this is the main reason for falling out of favor,
especially in the construction industry, a lead is the same as
the lag looked at from the opposite side. And overlap, on the
other hand, is a negative lag. So think about it for a second if you
say that activity number two starts three days before the
completion of activity number one, which means there's going to be an
overlap of three days between the durations of these two activities.
So as if we are moving in the opposite direction, therefore is
going to be a negative lag. Lag is usually going to be a positive
number. Overlap is going to be a negative number.
So when we look at the node diagram drawing, nodes should be
drawn as squares or as rectangles. Basically do not connect the nodes
from the top or the bottom. So this is wrong. And this is wrong
because again, we mentioned that the relationship is going to start
either from the start or the finish of the activity. So this is
somewhere in the middle, which does not mean mean anything,
especially that the boxes are of the same size, and the location of
the line does not represent any scale.
So connect size only. The left side represents the start side and
the right side represents the end or the finish side.
Left to right, we're going to start moving from right to left.
Instead of adding durations, we're going to subtract duration. Now
we're going to have another element, which is the element of
lags and overlaps. The lag, as we said, is going to be a positive
number, so it's going to be added as we move forward. And the lag,
which used to be a positive number added in the forward pass is going
to be subtracted in the backward pass. On the other hand, the
overlap, which we mentioned, is going to be a negative number, so
the overlap is going to be subtracted in the forward pass,
and when we reverse go in the backward pass, we're going to add
the overlaps again. Don't worry about that. We're going to see a
numerical example on network calculations, which is going to
illustrate this issue, and it's going to be extremely
straightforward. So the early start of an activity is equal to
the maximum of the early finishes of all of its predecessors, plus
any lag or overlap that's going to be affected in and the early
finish of an activity is equal to its early start plus its duration
in the backward backward pass
running from right to left. It's used to determine the late finish
and the late start of each activity. Late finish of the
activity is equal to the minimum of the early finishes of all of
its predecessors minus lag of overlap. Late start of the
activity is equal to late finish minus duration. Again, exactly the
same rules that we used for Arrow Diagramming Method.
Now again, we're going to be faced with the issues of floats, as we
learned last time we had total float and free float and we talked
about something called interfering float, which we're not going to
use. So here we're going to focus primarily on total float and free
float, the definition is exactly the same. The total float is the
amount of time by which a non critical activity can be delayed
without delaying the whole project. Whereas the free float,
we're going to change the last couple of words, it's the amount
of time by which a non critical activity can be delayed without
delaying its immediate successor. We're not looking to the end of
the project. We're just looking for the immediate successor. And
as we agreed last time, also the free float is a subset of the
total float, which means the free float can never exceed the total
float. Remember that quite well, because this is one of the common
mistakes that I usually see on assignments and on exams, someone
giving a total float of three and a free float of five that can
never happen. The maximum of the free float can be equal to the
total float of the activity. The minimum for the free float, it can
be equal to zero. The free float can never be a negative value
again. Remember that, because this is another common mistake. So the
free float has boundaries. The lowest one is zero, the highest
one is equal to the total float of the activity. Activities whose
total float is zero are on the critical path. And if an activity
is on the critical path, by default, its total float is zero,
and if the total float is zero, then definitely the free float is
also going to be equal to zero, because it cannot exceed the total
float, and it cannot be a negative value. Therefore the only value
left is zero.
To calculate the total float, the total float is equal to the late
finish minus the early finish, or the late start, minus the early
start. So again, you calculate it from either side of the activity,
late minus early for the same side, late start minus early
start, or late finish minus early finish.
Now for the free float, it's little bit more complex when it
comes to PDM, because we may have different types of relationships,
we may have lags and overlaps and so on and so forth. So I invented
a method to calculate the free float, and it's called the sad
method to calculate the free float. And if you follow it,
you're going to find that it's the simplest and easiest way to
calculate the free float of an activity.
The free float of an activity is equal to the total float of that
activity minus the largest total float of any of its immediate
successors. Listen again. It's equal to the total float of the
activity minus the largest of the total floats of any of its
immediate successors. So if we have an activity, having a total
float of five,
and its immediate successors have total floats of five, three and
six.
So in this case, according to the Saad method.
So the free float of this activity is equal to five minus the largest
total float of five, three and six, which is six, five minus six
is negative one. But we just said that the free float cannot be
negative. In case you get a negative value, put the free float
equal to zero. So another example, if an activity has two immediate
successors. The first activity has five days of total float. Its
immediate successors have floats of three and two. Then the free
float of this activity is equal to five minus the larger of the three
and two, which is three. So five minus three, that gives two days
of free float for that activity.
So there's a general
understanding on how to draw the activities in the network, and
it's something like this box here, where we divide it primarily into
seven components, seven compartments. In the middle here,
we're going to have the activity ID or its description. What is
that activity about? And in the middle, at the bottom, we're going
to have the duration, which is going to be given or calculated
primarily. And then, based on that, we're going to perform our
calculations. We're going to calculate the early start plus
duration gives the early finish, the late finish minus duration
gives the late start, late start, minus early start, or late finish,
minus early finish, is going to give the total float. So the only
number that we start with is the duration, and based on the
duration which we acquire from Q over P, remember that very simple
equation that we said is always going to be with us, Q divided by
P, Q, the amount of work to be done divided by P, the lowest
production rate of any of the resources involved in that
activity. So q over P gives the duration, and from that, we can
calculate all the other dates for that activity, depending on the
relationship linking this activity to other activities.
President's network, some people would would like to make a
distinction between
Aon and precedence. I do not usually make the distinction. I
consider them the same thing. So precedence network have these four
types of relationships, finish to start, start to start, finish to
finish and start to finish.
So besides relationship types, each relationship can be
accompanied by a lag or overlap value. So if it's a finish to
start with lag, meaning that the successor is going to start three
days after the completion of the predecessor. Examples for that. We
have placed the concrete for a slab,
and the next activity is to remove the four more for that slab. Of
course, we cannot remove the four more. As soon as we have finished
placing the concrete, we have to wait for this concrete to have
setting, initial setting, and part of the permanent setting
dependent, dependent on the code, depending on the span, depending
on the concrete mix, depending on the weather and other conditions
and so on. So we have to wait, for example, let's say a week after
the concrete has gained enough strength before being able to
remove the formal so in this case, we're going to say that removal of
four work is gonna lag seven days after the completion of placing
that concrete slab. The lag values indicate the amount of delay
between the two elements of the relationship described by the
relationship type. So if it's a start to start, the start of the
successor is gonna lag by certain number of days after the start of
the predecessor. If it's a finish to finish the completion of the
second activity is going to lag a certain number of days after the
completion of the first one, and so on and so forth.
Two ways to present different relationships. So if we have a
just an arrow spanning between the end of an activity and the start
of the other one. It would be a finish to start.
However, if we want to designate the start to start, we can draw it
from the start of the activity to the start of the successor, or use
the traditional finish to start designation and put on it the two
letters representing the connecting ends of the activity,
SS for start to start, or FF, for finish to finish. Honestly, I do
not prefer this however you might see it in some books or art or
articles or publications. I do not use this one. So forget about this
one. We're not gonna use it to designate start to start. This is
the one that we're going to use which has less confusion, it shows
that the relationship is from the start of the predecessor to the
start of the successor. And usually, in most cases here, we're
going to have a positive number, which represents a lag.
And similarly, for finish to finish is going to be from the.
Finish of one activity here to the finish of its immediate successor,
and so on.
Now, since we talked about
contiguous activities, which we meant that these are activities
that once started, cannot stop until the activity is complete,
the other option is called interruptible activities, which
are activities that can be interrupted, paused for a certain
number of days and then resumed at a later date and completed at the
end. So we have two different types of calculations. One of them
is going to be based on a contiguous activity assumption,
and the other one is going to be based on an interruptible activity
assumption. We're going to look at both examples in a numerical
example that we're going to solve in another lecture. So the
continuous the contiguous activities cannot be interrupted.
And once started, they must continue until they are finished,
for the finish to start relationship in the forward pass,
the early finish is equal to the early start plus the duration
and the early start of the successor activity is going to be
the maximum of all predecessor dates, which might be the largest
early finish of the previous activities, plus any lag or
overlap values between the two activities that's going to appear
on the arrow, or the relationship between the activities
in the backward pass. Again, the backward pass determines, or
provides the late dates, late start and late finish for the
activity. It begins at the last activity on the network moving
backwards, so the late finish of the activity is going to be the
minimum of all the successors, late start plus lag or overlap.
And the late start is going to be the late finish minus the
duration, because we move from right to left, subtract
in case of start to start again, it's exactly the same concept.
We're going to look at the largest number coming to the start of
this, this activity, whether it's coming coming from the immediate
predecessors, whether it has lags or overlaps and so on. We are
going to factor that into our calculation.
In some cases, we might have something called a dangling
activity or open ended activity. This is not a good thing to have
in a network, but it might happen in some projects. To give you an
example, something like landscaping activities. If we're
talking about a five story building, when can we start the
landscaping activities? Well, we can start them as soon as the
enclosure of the building is complete, so that we do not have
any heavy equipment on the outside of the building. We still have a
lot of work to be done inside the building, all the interior
finishing and so on and so forth, but we can start the landscaping
at that point. Now, when does the landscaping need to be finished?
It needs to be finished by the end of the project. If the landscape
is going to take two months, and the interior finishing is going to
take five months, if we start the landscaping, once the enclosure is
complete, it's going to start, and then it's going to be done in two
months. It still have, has three months of total float until the
end of the project, because we are still working on the inside. So in
this case, we say that this is an open ended activity, or dangling
activity.
A dangling activity has either no predecessors or no successors.
Should be avoided, as they reflect the false amount of float. Again,
it would appear in this case that the landscaping has a huge amount
of total float, but we can tie it to other activities on the
construction site. The dates may vary depending on whether the
activity duration is contiguous or interruptible. We're going to see
that again in the numerical example.
In case of finish to finish, the finish of one activity controls
the finish of another its successor, and it's used to relate
to activities that are done in parallel or may have lags. So
again, same calculations would apply a
a start to finish. As I mentioned before, it's very rarely used, and
it's usually linked linking one non construction activity, like a
marketing campaign, for example, to some construction activities.
It's used to identify activities whose starts are related to their
successors finish
difficult to identify a pair of such activities in construction.
Again, do not worry about that, because we're not going to have
many activities like that in in our calculations,
as we discussed before, one of the fatal loops.
In any network, one of the fatal errors in any network is the
existence of loops. A loop is activities going back and forth in
an unending fashion, so two or more activities linked in a
circular manner. You say that activity two is a successor to
activity one, and activity one is a successor to activity two. So it
keeps the calculations keep going in a circle, and that's a fatal
error. The software, if we use the software, is going to give you an
error that you cannot operate in a loop. It usually can be found in
relationships where the arrow turns backward. It's a fatal
mistake that should be avoided at all costs, and represents a faulty
logic. So this is basically a an introduction about presence
diagramming methods or
activity on node. We are going to see a in another lecture or
another example. We're going to see a numerical example on how to
draw them and how to make the calculations, and what's the mean
difference between ADM and PDM See you in another lecture. You.