EMÜ 222

Deterministic Operations Research

2023-2024 Spring

Week 10
Critical Path Method (CPM)
Program Evaluation and Review Technique (PERT)
Part 1

Winston W.L., Operations Research, Chapter 8

 Critical Path Method
• Network models can be used to schedule large
complex projects that consist of many
activities
• If the duration of activity is known with
certainty: Use critical path method (CPM)
• If the duration of activity is not known with
certainty: Use program evaluation and review
technique (PERT)
 Examples of Project Scheduling
• Scheduling construction projects (office
buildings, highways, swimming pools)
• Installing a new computer system
• Building a ship
• Designing and marketing a new product
 Project Network
• In order to analyze and schedule tasks of a
project, we will generate project networks
• In doing so, we need
– the specific tasks to be completed for doing the
project
– Predecessor/successor relationships among the
tasks
– Information regarding the duration of the tasks
 Project Network
• There could be two methods for representing
the project network
– AOA: Activity on Arc
– AON: Activity on Node

• Both methods are viable but your text book

prefers to use AOA representation
 AOA Representation
• Nodes: completion of one or more activities
• Arcs: Activities
Examples
A B Node 2: Completion of activity A
1 2 3
Here, A is the predecessor of B

1 A
C
3 Node 3: Completion of activities A & B
2 B
Here, A and B are predecessors of C
 AOA Representation
• Nodes: completion of one or more activities
• Arcs: Activities
Examples

B
A
1 2 Node 2: Completion of activity A
C
Here, A is the predecessor of both
B and C
 Rules in Constructing a Project Network
• Node 1 is the start of the project.
• There should be a finish node.
• Number the nodes in the network so that the node
representing the completion of an activity always has a
larger number than the node representing the beginning of
an activity.
• An activity should not be represented by more than one
arc in the network.
• Two nodes can be connected by at most one arc.

To avoid violating rules 4 and 5, we include dummy activities!

 Rules 4 and 5

Suppose A and B are two predecessors of C.

This representation violates Rule 5.

This representation is correct. Include a

dummy arc.
Node 3: Completion of B
Node 2: Completion of A and B
 Rules 4 and 5

Example
Activity Immediate Predecessor

A -
B -
C A
D A, B

A C
1 2 5

B D
3 4 6
 Example 1
Activity Immediate
Predecessor
A -
B A
C B
D B
E B
F E
G C, F

11
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A
1

12
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B
1 2

13
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B C
1 2 3

14
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B E
1 2 3

D
15
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B E F
1 2 3 4

D
16
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B E F G
1 2 3 4 5

D
17
 Activity Immediate Predecessor
A -
B A
C B
D B
E B
F E
G C, F

A B E F G
1 2 3 4 5 6

D
18
Example 2: Development of a Software Package

Activity Description Immediate Expected Duration

Predecessor (days)
A Perform market survey - 3
B Design graphic icons A 4
C Develop flowchart A 2
D Design input/output screens B, C 6
E Module 1 coding C 5
F Module 2 coding C 3
G Module 3 coding E 7
H Module 4 coding E, F 5
Merge modules and graphics
I D,G,H 8
and test program

19
 Activity Immediate
Predecessor
A -
B A
C A
D B, C
E C
F C
G E
H E, F
I D,G,H
B4 4 D6
A3
1 2
C2 G7 I8
E5 7 8
3 5

F3 H5
6
20
 Example 3
Activity Immediate Duration
Predecessor (Days)
A - 6
B - 9
C A,B 8
D A,B 7
E D 10
F C,E 12

21
 Activity Immediate Predecessor Duration (Days)
A - 6
B - 9
C A,B 8
D A,B 7
E D 10
F C,E 12

A6 3
C8
1 D7 F12
4 5 6
E10
B9
2

22
 Notation for Critical Path Method
• 𝒕𝒊𝒋 : Expected duration of activity (i,j)
• ET(i): Early Event Time of node i – The earliest
time at which the event corresponding to
node i can occur.
• LT(i): Late Event Time of node i – The latest
time at which the event corresponding to
node i can occur without delaying the
completion of the project.

23
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2
B - 9
3
C A,B 8
4
D A,B 7
5
E D 10
6
F C,E 12

24
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2 ET(1)+9 = 9
B - 9
3
C A,B 8
4
D A,B 7
5
E D 10
6
F C,E 12

25
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2 ET(1)+9 = 9
B - 9
3 max{ET(1)+6, ET(2)+0} = 9
C A,B 8
4
D A,B 7
5
E D 10
6
F C,E 12

26
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2 ET(1)+9 = 9
B - 9
3 max{ET(1)+6, ET(2)+0} = 9
C A,B 8
4 ET(3)+7 = 16
D A,B 7
5
E D 10
6
F C,E 12

27
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2 ET(1)+9 = 9
B - 9
3 max{ET(1)+6, ET(2)+0} = 9
C A,B 8
4 ET(3)+7 = 16
D A,B 7
5 max{ET(3)+8, ET(4)+10} = 26
E D 10
6
F C,E 12

28
 Example - ET
Activity Immediate Duration Node (i) ET(i)
Predecessor (days)
1 0
A - 6
2 ET(1)+9 = 9
B - 9
3 max{ET(1)+6, ET(2)+0} = 9
C A,B 8
4 ET(3)+7 = 16
D A,B 7
5 max{ET(3)+8, ET(4)+10} = 26
E D 10
6 ET(5)+12 = 38
F C,E 12

• ET(i) is the length of the longest path in the network from

node 1 to node i.
• The earliest project completion time is 38 days.
29
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1
A - 6
2
B - 9
3
C A,B 8
4
D A,B 7
5
E D 10
6 38
F C,E 12

30
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1
A - 6
2
B - 9
3
C A,B 8
4
D A,B 7
5 LT(6)-12 = 26
E D 10
6 38
F C,E 12

31
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1
A - 6
2
B - 9
3
C A,B 8
4 LT(5)-10 = 16
D A,B 7
5 LT(6)-12 = 26
E D 10
6 38
F C,E 12

32
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1
A - 6
2
B - 9
3 min{LT(4)-7),LT(5)-8} = 9
C A,B 8
4 LT(5)-10 = 16
D A,B 7
5 LT(6)-12 = 26
E D 10
6 38
F C,E 12

33
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1
A - 6
2 LT(3)-0 = 9
B - 9
3 min{LT(4)-7),LT(5)-8} = 9
C A,B 8
4 LT(5)-10 = 16
D A,B 7
5 LT(6)-12 = 26
E D 10
6 38
F C,E 12

34
 Example - LT
Activity Immediate Duration Node (i) LT(i)
Predecessor (days)
1 Min{LT(2)-9, LT(3)-6} = 0
A - 6
2 LT(3)-0 = 9
B - 9
3 min{LT(4)-7),LT(5)-8} = 9
C A,B 8
4 LT(5)-10 = 16
D A,B 7
5 LT(6)-12 = 26
E D 10
6 38
F C,E 12

35
 Notation for Critical Path Method
• TF(i,j): Total Float of activity (i,j) – The
maximum amount that the starting time of
that activity can be delayed beyond its earliest
possible start time without delaying the
project (assuming other activities are not
delayed) OR the amount that the duration of
the activity can be increased without delaying
the project

36
 Notation for Critical Path Method
𝑇𝐹 𝑖, 𝑗 = 𝐿𝑇 𝑗 − 𝐸𝑇 𝑖 − 𝑡𝑖𝑗
Node (i) ET(i) LT (i) Activity (i,j) TF(i,j)
1 0 0 A (1,3) 9-0-6 = 3
2 9 9 B (1,2) 9-0-9 = 0
3 9 9 C (3,5) 26-9-8 = 9
4 16 16 D (3,4) 16-9-7 = 0
5 26 26 E (4,5) 26-16-10 = 0
6 38 38 F (5,6) 38-26-12 = 0
Dummy (2,3) 9-9-0 = 0

37
 Finding a Critical Path
- If an activity has a total float of 0, then any delay in
the start of the activity (or the duration of the activity)
will delay the completion time of the project. This
activity is called as critical activity.
- Delaying the start time of a critical activity by ∆ days/
increasing the process time of a critical activity by ∆
days delays the project ∆ days.
- A path from node 1 to the finish node that consists
entirely of critical activities is called a critical path.

38
 Finding a Critical Path
What is the critical path of Example 3?

TF = 0 for B, D, E,
3 F, Dummy
1 D7 F12
4 5 6
E10 Critical activities
B9
2

The length of the critical path gives the project completion

time. It is the longest path from start to finish.
38 is the project completion time.
39
 Notation for Critical Path Method
• FF(i,j): Free Float of activity (i,j) – The amount by
which the starting time of the activity corresponding
to arc (i,j) (or the duration of the activity) can be
delayed without delaying the start of any later
activity beyond its earliest possible time.

𝐹𝐹 𝑖, 𝑗 = 𝐸𝑇 𝑗 − 𝐸𝑇 𝑖 − 𝑡𝑖𝑗

40
 Notation for Critical Path Method
𝐹𝐹 𝑖, 𝑗 = 𝐸𝑇 𝑗 − 𝐸𝑇 𝑖 − 𝑡𝑖𝑗
Node (i) ET(i) LT (i) Activity (i,j) FF(i,j)
1 0 0 A (1,3) 9-0-6 = 3
2 9 9 B (1,2) 9-0-9 = 0
3 9 9 C (3,5) 26-9-8 = 9
4 16 16 D (3,4) 16-9-7 = 0
5 26 26 E (4,5) 26-16-10 = 0
6 38 38 F (5,6) 38-26-12 = 0
Dummy (2,3) 9-9-0 = 0

The relation
FF(i,j) ≤ TF(i,j)
between FF and TF: 41
 Another Example
C4
A6 3
5 F4
1 D7 G12
4 6 7
E10
B9
2
Activity (i,j) TF(i,j)
Node (i) ET(i) LT (i) A (1,3) 9-0-6 = 3
1 0 0 B (1,2) 9-0-9 = 0
2 9 9 C (3,5) 22-9-4 = 9
3 9 9 D (3,4) 16-9-7 = 0
4 16 16 E (4,6) 26-16-10 = 0
5 13 22 F (5,6) 26-13-4 = 9
6 26 26 G (6,7) 38-26-12=0
7 38 38 Dummy (2,3) 9-9-0 = 0 42
 Another Example
C4
A6 3
5 F4
1 D7 G12
4 6 7
E10
B9
2

Activity (i,j) TF(i,j) FF(i,j)

Node (i) ET(i) LT (i) A (1,3) 9-0-6 = 3 9-0-6 = 3
1 0 0 B (1,2) 9-0-9 = 0 9-0-9 = 0
2 9 9 C (3,5) 22-9-4 = 9 13-9-4 = 0
3 9 9 D (3,4) 16-9-7 = 0 16-9-7 = 0
4 16 16 E (4,6) 26-16-10 = 0 26-16-10 = 0
5 13 22 F (5,6) 26-13-4 = 9 26-13-4 = 9
6 26 26 G (6,7) 38-26-12=0 38-26-12=0
7 38 38 Dummy (2,3) 9-9-0 = 0 9-9-0 = 0 43
 Using LP to Find the Critical Path

𝑥𝑗 : The time that the event of node 𝑗 occurs

𝑀𝑖𝑛 𝑧 = 𝑥6 − 𝑥1
s.t.
𝑥2 ≥ 𝑥1 + 9 Optimal solution:
𝑧 = 38, 𝑥1 = 0, 𝑥2 = 9,
𝑥3 ≥ 𝑥1 + 6
𝑥3 = 9, 𝑥4 = 16,
𝑥3 ≥ 𝑥2 + 0 𝑥5 = 26, 𝑥6 = 38
𝑥4 ≥ 𝑥3 + 7
Check the Excel solution.
𝑥5 ≥ 𝑥3 + 8
𝑥5 ≥ 𝑥4 + 10
𝑥6 ≥ 𝑥5 + 12
𝑥𝑗 𝑈𝑅𝑆 44
 Using LP to Find the Critical Path
Intrepreting the Sensitivity Report of Excel
- If a constraint has 0 as the shadow price, delaying the
activity (corresponding to that constraint) will not
affect z (the project completion time).

- If a constraint has 1 as the shadow price, delaying this

activity by 1 day increases project completion time by 1
day , so that it is a critical activity.

45
 Using LP to Find the Critical Path
Intrepreting the Sensitivity Report of Excel
Variable Cells
Final Reduced Objective Allowable Allowable
Cell Name Value Cost Coefficient Increase Decrease
$B$3 x(1) 0 0 -1 1E+30 0
$B$4 x(2) 9 0 0 1E+30 0
$B$5 x(3) 9 0 0 1E+30 0
$B$6 x(4) 16 0 0 1E+30 0
$B$7 x(5) 26 0 0 1E+30 0
$B$8 x(6) 38 0 1 1E+30 0

Constraints
Final Shadow Constraint Allowable Allowable
Cell Name Value Price R.H. Side Increase Decrease
B (1,2) 9 1 9 1E+30 3
A (1,3) 9 0 6 3 1E+30
Dummy (2,3) 0 1 0 1E+30 3
D (3,4) 7 1 7 1E+30 9
C (3,5) 17 0 8 9 1E+30
E (4,5) 10 1 10 1E+30 9
F (5,6) 12 1 12 1E+30 38
46