Solution To Building A House: 1 4 3 A 5 B 8 C 10 E 4 G 3 F 6
Solution To Building A House: 1 4 3 A 5 B 8 C 10 E 4 G 3 F 6
Consider the (simplified) list of activities and predecessors that are involved in
building a house, as listed in Table 1.
a. Draw a project network and use LP to find the critical path and the
minimum number of days needed to build the house.
Here is one possible network diagram (an activity-on-arc diagram). The arrows
(arcs) represent activities; the balls (nodes) have been numbered for ease of
discussion:
F
1 3 4
6
A G
5 E 3
4
0 C 5
B
Start 8 10 End
D
5
2
1
Based on 5-45 (p. 259) in Practical Management Science (2nd ed., Winston and Albright, 2001
Duxbury Press). Solution by David Juran, 2002.
Managerial Formulation
Decision Variables
We are trying to decide when to begin and end each of the activities.
Objective
Minimize the total time to complete the project.
Constraints
Each activity has a fixed duration.
There are precedence relationships among the activities.
We cannot go backwards in time.
Mathematical Formulation
This is an example of a type of project-scheduling problem that can be solved
with the Critical Path Method (CPM). These problems can be solved by hand, or
with any of a large number of project management software packages. They can
also be solved using linear programming, as is demonstrated here.
Decision Variables
Define the nodes to be discrete events. In other words, they occur at one exact
point in time. Our decision variables will be these points in time.
Define ti to be the time at which node i occurs, and at which time all activities
preceding node i have been completed.
Define t0 to be zero.
Objective
Minimize t5.
Constraints
There is really one basic type of constraint. For each activity x, let the time of its
starting node be represented by tjx and the time of its ending node be represented
by tkx. Let the duration of activity x be represented as dx.
For every activity x, t kx t jx d x
For every node i, t i 0
It is also possible to identify the critical path by looking at the Solver answer
report:
Constraints
Cell Name Cell Value Formula Status Slack
$H$12 A 5 $H$12>=$J$12 Binding 0
$H$13 B 8 $H$13>=$J$13 Binding 0
$H$14 C 10 $H$14>=$J$14 Binding 0
$H$15 D 13 $H$15>=$J$15 Not Binding 8
$H$16 E 4 $H$16>=$J$16 Binding 0
$H$17 F 6 $H$17>=$J$17 Binding 0
$H$18 G 3 $H$18>=$J$18 Binding 0
Conclusions
The project will take 26 days to complete.
The only activity that is not critical is the electrical wiring.
Managerial Formulation
Decision Variables
Now the problem is not only when to schedule the activities, but also which
activities to accelerate. (In CPM jargon, accelerating an activity at an additional
cost is called “crashing”.)
Objective
Minimize the total cost of crashing.
Constraints
The project must be finished in 20 days.
Each activity has a maximum amount of crash time.
Each activity has a “basic” duration. (These durations were considered to have
been fixed in Part a; now they can be reduced.)
There are precedence relationships among the activities.
We cannot go backwards in time.
Mathematical Formulation
Decision Variables
Define the number of days that activity x is crashed to be Rx. For each activity
there is a maximum number of crash days Rmax, x
Define the crash cost per day for activity x to be Cx
Constraints
For every activity x, t kx t jx d x R x
For every activity x, R x R max, x
For every node i, t i 0
Solution Methodology
Here’s the spreadsheet model:
A B C D E F G H I J K L M N O
F
1
1 3 4
2 6
3 30
4 A G
5 t0 t1 t2 t3 t4 t5
6 0 1 51 1 1 1 E 3
7 4
8 0 B
20 <-- Max Completion Time C 5
9
10Start 8 10 End
11 t0 t1 t2 t3 t4 t5 Improved Duration Basic Duration Crash Time Max Crash Cost/Time
12 A -1 1 0 0 0 0 1 >= 4 5 1 2 $ 30
13
14
B
C
0
0
-1
0
1
-1
0
0
0
1
0
0
0
0
>=
>=
8
10
D 8
10
0
0
3
1
$
$
15
20
15 D 0 0 -1 0 0 1 0 >= 5 5 5 0 2 $ 40
16 E 0 0 -1 1 0 0 0 >= 4 4 0 2 $ 20
17 F 0 0 0 -1 1 0 2
0 >= 6 6 0 3 $ 30
18 G 0 0 0 0 -1 1 0 >= 3 3 0 1 $ 40
1 3 F 4
6
A G
3 E 3
3
0 C 5
Start B End
5 9
D
5
2