Title: Samuel Labi and Fred Moavenzadeh
1 Probabilistic Planning Part I
1.040/1.401 Project Management
March 19, 2007
- Samuel Labi and Fred Moavenzadeh
- Department of Civil and Environmental Engineering
- Massachusetts Institute of Technology
2 Recall
3Outline for this Lecture
- Deterministic systems decision-making the
general picture - Probabilistic systems decision-making the
general picture - the special case of project planning
- Why planning is never deterministic
- Simplified examples of deterministic,
probabilistic planning - - everyday life
- - project planning
- Illustration of probabilistic project planning
- PERT Basics, terminology, advantages,
disadvantages, example
4 Deterministic Systems Planning and Decision-making
Influence of deterministic inputs on the outputs
of engineering systems -- the general picture
FIXED VALUES of the Outputs (system performance
criteria, etc.)
Input variables and their FIXED VALUES
Final evaluation result and decision
Deterministic Analysis of Engineering System
Examples queuing systems, network systems, etc.
Alt. 1
Alt. k
Alt. n
A SINGLE evaluation outcome
Examples costs, time duration, quality, interest
rates, etc.
VALUE OF combined output (index, utility or
value) representing multiple performance measures
of the system
5But engineering systems are never
deterministic!Why?
- For sample, for Project Planning Systems
- Variations in planning input parameters
- Beyond control of project manager
- Categories of the variation factors
- Natural (weather -- good and bad, natural
disasters, etc.) - Man-made (equipment breakdowns, strikes, new
technology, worker morale, poor design, site
problems, interest rates, etc.) - Combined effect of input factor variation is a
variation in the outputs (costs, time, quality)
6Why planning is never deterministic -- II
Influence of stochastic inputs on the outputs of
engineering systems -- the general picture
Input variables and their probability
distributions
Outputs (system performance criteria, etc.) and
their probability distributions
Probability distributions for individual outputs
f(O1)
fX
Variability of final evaluation result and
decision
f(O2)
fY
Probabilistic Analysis of Engineering System
f(OM)
PAlt. i
Examples queuing systems, network systems, etc.
Alt. 1
Alt. 2
Alt. n
fZ
f(OCOMBO)
Discrete probability distribution for evaluation
outcome PAlt. i is the probability that an
alternative turns out to be most desirable or
most critical
Examples costs, time durations, quality,
interest rates, etc.
Probability distribution for combined output
(index, utility or value) representing multiple
performance measures
7 Why planning is never deterministic -- III
Probability distribution? What exactly do you
mean by that?
Input variable and its probability distribution
f
See survey for dinner durations (times) of class
members
8 Probability Distribution for your Dinner Times
9 Probability that a randomly selected student
spends less than T minutes for dinner is less
than T P(T lt T)
10 P(T)
Mean µ Std dev s
T
µ
0
T
Standard Normal Transformation
Probability that a randomly selected student
spends less than T minutes for dinner is less
than T1 P(T lt T)
f(Z)
Mean 0 Std dev 1
Z
Z
µ 0
11 P(T)
Mean µ Std dev s
T
µ
0
T
Standard Normal Transformation
Probability that a randomly selected student
spends less than T1 P(T lt T)
more than T1 P(T gt T)
f(Z)
Mean 0 Std dev 1
Z
Z
µ 0
12 P(T)
Mean µ Std dev s
T
µ
0
T
Standard Normal Transformation
Probability that a randomly selected student
spends less than T P(T lt T)
more than T P(T gt T) between T1 and T2
P(T1 lt T lt T2)
f(Z)
Mean 0 Std dev 1
Z
Z
µ 0
13 Probability is the area under the probability
distribution/density curves) Probability can be
found using any one of three ways - coordinate
geometry - calculus - statistical tables
14 Like-wise, we can build probability distributions
for project planning parameters by - using
historical data from past projects, OR -
computer simulation
And thus we can find the probability that project
durations falls within a certain specified range
15 - Probabilistic planning of project management
systems can involve uncertainties in - Need for an Activity (need vs. no need)
- Durations
- Activity durations
- Activity start-times and end-times
- Cost of activities
- Quality of Workmanship and materials
- Etc.
16 - Probabilistic planning of project management
systems can involve uncertainties in - Need for an Activity (need vs. no need)
- Durations
- Activity durations
- Activity start-times and end-times
- Cost of activities
- Quality of Workmanship and materials
- Etc.
17 Influence of stochastic inputs -- the specific
picture of project planning
Input variable and its probability distribution
Outputs (system performance criteria, etc.) and
their probability distributions
Probability distributions for the output
Variability of final evaluation result and
decision
Probabilistic Analysis of Project Planning
PAlt. i
fX
f(O1)
Project planning system
Path 1
Path 2
Path n
In this case, Output is the identification of
the critical path for the project.
Frequency distribution or probability
distribution for evaluation outcome PAlt.i is
the probability that a given path turns out to be
the critical path
Project activity durations
18 Probabilistic planning An example in everyday
life
KEY
Activity
Perfectly deterministic
Start Time
Activity Duration
Finish Time
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
START 8AM
Duration 5hrs
FINISH 1 PM
START 7AM
Duration 1hr
FINISH 8AM
US Meeting in Class For this Lecture
FINISH 230
START 1PM
Duration1.5hr
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
START 8AM
Duration 5 hr
FINISH1 PM
START 7 AM
Duration 1 hr
FINISH 8AM
START 1 PM
Duration1.5hr
FINISH 230
19 Probabilistic planning An example in everyday
life
KEY
Activity
Partly deterministic, Partly probabilistic
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
20 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
21 Probabilistic planning An example in everyday
life
KEY
Activity
Partly deterministic, Partly probabilistic
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
22 Probabilistic planning An example in everyday
life
KEY
Activity
Partly deterministic, Partly probabilistic
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
23 Probabilistic planning An example in everyday
life
KEY
Activity
Partly deterministic, Partly probabilistic
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
24 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
25 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
26 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
Deterministic Parts
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
27 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
28 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
ES 1255
Duration 1.5 hrs
EF 225
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
29 Probabilistic planning An example in everyday
life
KEY
Partly deterministic, Partly probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration 5 hrs
EF NOON
ES 6AM
Duration 1 hour
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
Probabilistic Parts
ES 1255
Duration 1.5 hrs
EF 225
LS 1 PM
LF 230
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration 5 hour
EF 1245
ES 645
Duration 1 hr
EF 745
ES 6AM
Duration 1 hour
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
30 Probabilistic planning An example in everyday
life
KEY
Fully probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration
EF NOON
ES 6AM
Duration
EF 7AM
US Meeting in Class For this Lecture
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
µ 5 hr s 0.6hr
µ 1hr s 0.25hr
ES 1255
Duration
EF 225
LS 1 PM
LF 230
µ 1.5 hr s 0.31hr
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 745
Duration
EF 1245
ES 645
Duration
EF 745
ES 6AM
Duration
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
µ 5 hr s 0.35 hr
µ 1 hr s 0.15 hr
µ 1hr s 0.2 hr
31 Probabilistic planning An example in everyday
life
KEY
Fully probabilistic
Activity
Earliest Start
Duration of Activity
Earliest Finish
Latest Start
Latest Finish
SAM Bathing, Breakfast,, Reading, Transport to
MIT, etc.
SAM Waking up and meditating
ES 7AM
Duration
EF NOON
ES 6AM
Duration
EF 7AM
US Meeting in Class For this Lecture
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
µ 5 hr s 0.6hr
µ 1hr s 0.25hr
ES 1255
Duration
EF 225
LS 1 PM
LF 230
µ 1.5 hr s 0.31hr
YOU Preparing for Classes, etc.
YOU showing up at Other Classes
YOU missing the lecture
ES 755
Duration
EF 1255
ES 655
Duration
EF 755
ES 6AM
Duration
EF 7AM
LS 7AM
LF 8AM
LS 8AM
LF 1 PM
LS 7AM
LF 8AM
µ 5 hr s 0.35 hr
µ 1 hr s 0.15 hr
µ 1hr s 0.2 hr
32Probabilistic planning An example in Project
Management
KEY
Activity Name
Early Start
Early Finish
Duration
Late Start
Late Finish
Activity A
Month 4
Month 8
4 months
Month 15
Month 11
Activity R
Activity G
Activity J
Activity W
Month 0
Month 4
Month 4
3 months
Month 7
Month 10
5 months
Month 15
4 months
Month 15
2 months
Month 17
Month 4
Month 0
Month 10
Month 7
Month 15
Month 10
Month 17
Month 15
Activity C
Activity M
Month 4
Month 10
Month 10
Month 13
6 months
3 months
Month 10
Month 4
Month 15
Month 12
33 Probabilistic planning An example in Project
Management
KEY
Activity Name
Early Start
Early Finish
Duration (O-M-P)
Late Start
Late Finish
O Optimistic (earliest time) M Most probable
time P Pessimistic (latest time)
Activity A
Month 4
4 Months (3-4-5)
Month 8
Month 15
Month 11
Activity R
Activity G
Activity J
Activity W
Month 0
4 Months (3-4-8)
Month 4
Month 4
3 Months (2-3-5)
Month 7
Month 10
5 Months (2-5-6)
Month 15
Month 15
2 Months (1-2-4)
Month 17
Month 4
Month 0
Month 10
Month 7
Month 15
Month 10
Month 17
Month 15
Activity C
Activity M
Month 4
6 Months (3-6-7)
Month 10
Month 10
3 Months (1-3-5)
Month 13
Month 10
Month 4
Month 15
Month 12
34 Probabilistic planning involving activity
durations An Illustration
Activity A
Month 4
4 Months (3-4-5)
Month 8
Month 15
Month 11
Activity R
Activity G
Activity J
Activity W
Month 0
4 Months (3-4-8)
Month 4
Month 4
3 Months (2-3-5)
Month 7
Month 10
5 Months (2-5-6)
Month 15
Month 15
2 Months (1-2-4)
Month 17
Month 4
Month 0
Month 10
Month 7
Month 15
Month 10
Month 17
Month 15
Activity C
Activity M
Month 4
6 Months (3-6-7)
Month 10
Month 10
3 Months (1-3-5)
Month 13
Month 10
Month 4
Month 15
Month 12
35 Probabilistic planning involving activity
durations An Illustration
Activity A
Month 4
4 Months (3-4-5)
Month 8
Month 15
Month 11
Activity R
Activity G
Activity J
Activity W
Month 0
4 Months (3-4-8)
Month 4
Month 4
3 Months (2-3-5)
Month 7
Month 10
5 Months (2-5-6)
Month 15
Month 15
2 Months (1-2-4)
Month 17
Month 4
Month 0
Month 10
Month 7
Month 15
Month 10
Month 17
Month 15
Activity C
Activity M
Month 4
6 Months (3-6-7)
Month 10
Month 10
3 Months (1-3-5)
Month 13
Month 10
Month 4
Month 15
Month 12
- Lets say we have lots of data on the durations
of each activity. Such data is typically from - Historical records (previous projects)
- Computer simulation (Monte Carlo)
36 Probabilistic planning involving activity
durations An Illustration
Activity A
Month 4
4 Months (3-4-5)
Month 8
Month 15
Month 11
Activity R
Activity G
Activity J
Activity W
Month 0
4 Months (3-4-8)
Month 4
Month 4
3 Months (2-3-5)
Month 7
Month 10
5 Months (2-5-6)
Month 15
Month 15
2 Months (1-2-4)
Month 17
Month 4
Month 0
Month 10
Month 7
Month 15
Month 10
Month 17
Month 15
Activity C
Activity M
Month 4
6 Months (3-6-7)
Month 10
Month 10
3 Months (1-3-5)
Month 13
Month 10
Month 4
Month 15
Month 12
37 - Calculate the Expected Duration of each path,
and Identify the Critical Path on the basis
of the mean only
38 - Calculate the Expected Duration of each path,
and Identify the Critical Path on the basis
of both the mean and the std dev
Path R-C-J-W
Path R-C-M-W
Path R-G-J-W
Path R-A-W
39 Critical Path
Path R-C-J-W
Path R-C-M-W
Path R-G-J-W
Path R-A-W
40 Critical path here can be considered as that
with - Longest duration (mean) - Greatest
variation (stdev)
41 Consider the following hypothetical project paths
Path P-H-W-R
Path P-H-M-R
Path P-Q-R
Path P-F-W-R
On the basis of mean duration only, Path P-H-W-R
is the critical path
On the basis of the variance of durations only,
Path P-F-W-R is the critical path
How would you decide the critical path on the
basis of both mean duration and variance of
durations?
42 - Better way to identify critical path is using the
amount of slack in each path (see later slides)
43 - Another Example of Probabilistic Project
Scheduling - - Monte Carlo Simulation
- - Similar activity structure as before, but
Start and End activities are dummies (zero
durations). - See Excel Sheet Attached
44 Benefits of Probabilistic Project Planning
Discussed in previous slides
- Helps identify likely critical paths in
situations where there is great uncertainty - Helps ascertain the likelihood (probability) that
overall project duration will fall within a given
range - Helps establish a scale of criticality among
the project activities
45 Benefits of Probabilistic Project Planning
- Helps identify likely critical paths in
situations where there is great uncertainty - Helps ascertain the likelihood (probability) that
overall project duration will fall within a given
range - Helps establish a scale of criticality among
the project activities
Discussed in subsequent slides
46 - Probabilistic planning
- Is it ever used in real-life project management?
47A Tool for Stochastic Planning PERT
- Program Evaluation and Review Technique (PERT)
- - Need for PERT arose during the Space Race,
in the late fifties - - Developed by Booz-Allen Hamilton for US Navy,
and Lockheed Corporation - - Polaris Missile/Submarine Project
- - RD Projects
- - Time Oriented
- - Probabilistic Times
- - Assumes that activity durations are Beta
distributed
48PERT Parameters
- Optimistic duration a
- Most Likely duration m
- Pessimistic duration b
- Expected duration
- Standard deviation
- Variance
49Steps in PERT Analysis
- Obtain a, m and b for each activity
- Compute Expected Activity Duration dte
- Compute Variance vs2
- Compute Expected Project Duration DTe
- Compute Project Variance VS2 as Sum of Critical
Path Activity Variance - In Case of Multiple Critical Path Use the One
with the Largest Variance - Calculate Probability of Completing the Project
50PERT Example
51PERT ExampleFinding the Standard Deviation of
the duration of a given path comprising
Activities C, E, and G.
52PERT AnalysisFinding probability that project
duration is less than some valueExample
probability that project ends before 10 months
53Probability that the project will end before 13
months
54Probability that the project will have a duration
between 9 and 11.5 months
55PERT Advantages
- Includes Variance
- Assessment of Probability of Achieving a Goal
56PERT Disadvantages
- Data intensive - Very Time Consuming
- Validity of Beta Distribution for Activity
Durations - Only one Critical Path considered
- Assumes independence between activity durations
57PERT Monte Carlo Simulation
- Determine the Criticality Index of an Activity
- Used 10,000 Simulations, Now from 1000 to 400
Have Been Reported as Giving Good Results
58PERT Monte Carlo Simulation Process
- Set the Duration Distribution for Each Activity
- Generate Random Duration from Distribution
- Determine Critical Path and Duration Using CPM
- Record Results
59Example Network
D 2.83
C
A 2.17
A
G 2
D
F 4
B 6
F
B
E 5.17
G
End
End
Start
Start
C 3.83
E
60Monte Carlo Simulation Example
61Monte Carlo Simulation Example
Activity Duration
62Project Duration Distribution
10
9
8
7
6
Frequency
5
4
3
2
1
0
Project Length
29
17
23
24
25
26
27
28
18
19
20
21
22
63Probability Computations from Monte Carlo results
Number of Times Project Finished in Less Than or
Equal to T units Total Number of Replications
Number of Times Project Finished in More Than or
Equal to T units Total Number of Replications
ETC.
64Criticality Index for an Activity
Definition Proportion of Runs in which the
Activity is in the Critical Path
65 Criticality Index for a Path
Definition I (Naïve Definition) Proportion of
Runs in which the Activity is in the Critical
Path (see Slide 60)
66 Criticality Index for a Path
Definition I (Naïve Definition) Proportion of
Runs in which the Activity is in the Critical
Path (see Slide 60)
67 Criticality Index for a Path
Definition I Proportion of Runs in which the
Activity is in the Critical Path (see Slide 60)
68 Criticality Index for a Path
Definition II How much slack exists in that
path. Less Slack Higher criticality More
Slack Lower criticality
Ranges from 0 to 100
minimum total float maximum total
float total float or slack in current path
Using the index, we can rank project paths from
most critical to least critical
69 Criticality Index for a Path
Definition II See Example In Excel File (Path
Criticality Slide)
70References
- Kerzner, Harold, Project Management A Systems
Approach to Planning, Scheduling, and
Controlling, John Wiley and Sons, New York,
2000,7th Edition. - Meredith, Jack, R., and Samuel J. Mantel, Jr.,
Project Management A Managerial Approach, John
Wiley and Sons, New York, 2000, 4th Edition. - Halpin, Daniel, W., and Woodhead, Ronald, W.,
Construction Management, John Wiley and Sons,
New York, 2000, 2nd Edition. - Schtub, Avraham, Bard, Johnson and Globerson,
Schlomo. Project Management Engineering
Technology, and Implementation, Prentice Hall,
New Jersey, 1994. - Callahan, Quackenbush, and Rowings. Construction
Project Scheduling, McGraw-Hill, New York 1992. - Pritsker, A. Alan, C. Elliot Sigal, Management
decision Making, Prentice Hall, Englewood
Cliffs, NJ, 1983. - Thamhain, Hans, Engineering Management, John
Wiley and Sons, New York, 1992. - Software Designed for Project Control Class
1.432, MIT by Roger Evje, Dan Lindholm, S. Rony
Mukhopadhyay, Chun-Yi Wang. Fall Semester Term
Project, 1996. - Albano (1992), An Axiomatic Approach to
Performance-Based Design, Doctoral Thesis, MIT - Eppinger, S.D (1994). A Model-Based Method for
Organizing Tasks in Product Development. Research
in Engineering Design 6 1-13, Springer-Verlag
London Limited. - Eppinger, S.D. (1997) Three Concurrent
Engineering Problems in Product Development
Seminar, MIT Sloan School of Management. - Moder, Phillips and Davis (1983). Project
Management with CPM, PERT and Precedence
Diagramming, Third Edition. Van Nostrand Reinhold
Company. - Naylor, T (1995). Construction Project
Management Planning and Scheduling. Delmar
Publishers, New York. - Suh N.P. (1990). The Principle of Design. Oxford
University Press, New York. - Badiru and Pulat (1995) Comprehensive Project
Management, Prentice Hall, New Jersey. - Hendrickson and Au (1989) Project Management for
Construction, Prentice Hall, New Jersey. - Pritsker, Sigal, and Hammesfahr (1994). SLAM II
Network Models for Decision Support - Taylor and Moore (1980). RD Project Planning
with Q-GERT Network Modeling and Simulation,
Institute of Management Sciences. - http//www.enr.com/