You are on page 1of 23

# PERT Program Evaluation & Review Technique

## PERT(Program Evaluation and Review Technique)

A management tool that is suited for research and

development projects which are generally nonrepetitive in nature and is sometimes called once-through projects. Developed in the late 1950s for the US Navys Polaris Missile program Duration time for activities in research and development projects are subject to uncertainty

PERT
PERT handles the uncertainty problem by the use

## of three time estimates, namely:

The optimistic time
The most likely or probabilistic time The pessimistic time

## the same, techniques were developed independently Difference:

CPM uses a fixed duration for each activity
PERT uses a probability distribution

## Assumptions when using PERT

Each activitys duration can be represented by a continuous probability distribution whose mean (te), standard deviation, te variance (v) can be calculated 2. The distribution of the critical paths duration can be found from the tes and vs of the activities on the critical path.
1.

## Sequence of events for PERT

1.

2.

3.

4.
5.

Plan activities and logic as in CPM, but instead of assigning a fixed duration, assign three duration estimates: optimistic, pessimistic, and the most likely. Translate the three estimates into a continuous distribution; then find its mean (te), standard deviation, and variance Use each activitys te as its duration, and find the critical path Combine activities on the critical path to get a probabiliy distribution for the critical path Using the critical paths distribution, make inferences about the likelihood of the project being completed on or after any given day

## How are the three estimates determined?

If we have a history on an activity : Optimistic estimate best time ever
The estimated minimum time an activity will take
Expected activity duration when all the breaks are

## right Its probability is about 1 in 100 or 1%

Pessimistic estimate worse time ever
Estimated maximum time that will be required under

## How are the three estimates determined?

If we have a history on an activity : Most likely or probabilistic time an average of all previous times
It is the estimated normal length of time an activity

takes Time expected that will most frequently be required for an activity if it were done again and again under identical conditions

## How do we describe an activitys continuous probability distribution from our estimates?

We describe it by its
Mean (te),

Standard deviation te
variance (v)

Calculated as: a 4m b te
te
6 ba 6

te 2

Where:

a = optimistic m = most likely b = pessimistic Then, find the critical path by using te of each activity as

How do we combine the distributions of the critical path activities to find the distribution of the critical path?
The mean of the critical path TE is the sum of the

means te of the activities on the critical path, and the variance of the critical path V is the sum of the variances of the same activities. The standard deviation of the critical path te =V

How do we make inferences about the total projects duration once we have described the distribution of the critical path?
Assumption: the critical path does not change Statistics: Critical path has a normal distribution by using

the central limit theorem The probability of completing an activity for any specific time, TS can be determined by using the Z-table.
Z TS T E

TE

Where: mean

Z = number of standard deviations from TE = critical path (project) mean te = critical path (project) standard deviation TS = any date you choose

E.g.
The critical path had a TE = 20 and a te = 2 The probability of finishing by the end of day 20 is:
20 20 0 2 From Z table: Z = 0; therefore, P = 50% Z

## The probability of finishing by the end of day 18 is: 18 20 Z 1 2

From the Z table, Z=-1, therefore P=16 percent.

Note: for any day, Ts, the Z table gives the probability of

finishing on or before the end of that day, not only on that day.

## Terms and subscripts

By the end of the day any time from the past

until the end of the day, subscipt -30 On a day itself, subcript 30 After the end of a day, subscript 30-

Example:
Assume we perform the CPM calculations with

## durations = te, and we find the critical activities given.

Find the following: 1. TE 2. te 3. the probability that the project will finish by the end of day 31 4. the probability that the project will finish before the start of day 37

Example:
5. the probability that the project will finish during day 34 6. the probability that the project will finish on days 34,

## 35 or 36 7. the date of completion with at least 93 percent confidence

Solution:

Solution:

Solution:

Solution:

Solution:

percen t

Solution:

percen t

Solution:
In the preceding example, we assumed that there was only one path that could ever be critical. If there is one more possible critical path, we must use joint probabilities. Looking at the network with three possible critical paths:

Exercises:
Fill in the given tables and answer the question. Note round te, te , v, and Te to two decimal places, and round TE to the nearest whole day.

Find the following: 1. TE 2. te 3. Find the probability that the project will finish by the end of day 26 4. Find the probability that the project will finish before the start of day 25 5. Find the probability that the project will finish during day 24

6.