You are on page 1of 9

Project Selection ( Lesson 5 )

Project selection is a sequel to Project identification and Project Appraisal and is the First stage of Project Life Cycle Multiple Criteria in Projects : Investment -Future growth prospects Return -Similarity existing business Pay back -Environmental implications NPV Risk Evaluation of Projects : Point to note is, very rarely would one project emerge as the best fulfilling all the criteria listed for the project. However, if this happens, then it is called the DOMINANT project and it must be chosen. In practice there would be a set of Non-dominant projects and we may have to chose one amongst them depending upon the Organizational constraints. The selection will become a complex process because, some of them may be excellent on one criteria ( say rate of return ), but bad on other ( say Risk factor ) This conflict is resolved by Management decision as to what they are looking for ?

criteria Projects
1

C1 C

Thus project P1 o gets Xm2 and so


Important Observations in this Problem : 1. We notice clearly that P1 is better than P2 for both NPV & Return since it is higher compared to P2. 2. Similarly Project P6 is higher in both criteria compared to projects P2, P3, P4 & P5 ( except for P7 & P8, which are higher than P6 in NPV ) 3. Similarly P8 dominates P4, P5, P7 for both criteria. 4. Hence we see that from the set of these 8 Projects, P1, P6 and P8 are dominating all the other projects and hence we can say these three projects are NON DOMINANT SOLUTIONS amongst the set of 8 projects. 5. If we were to choose the projects we would obviously go for P1, P6 or P8, the NON DOMINANT SOLUTIONS, which is also known as PARETO OPTIMAL SOLUTION. 6. Between P1 and P6 we notice that while P1 gives higher RETURN, its NPV is lower, while P6 gives Higher NPV, but lower RETURN. Hence if we need to choose the best out of the two, then the choice is what we are looking for, better NPV or Higher Return, and accordingly choose either P6 or P1 respectively. The Preferred Solution It could be the Dominant Solution, if it exists OR It could be the Non-Dominant Solution or Pareto Optimal Solution set of projects OR It could be the solution from out of the non-dominant solutions which will involve trade off that will be governed by the priorities or weightings attached by the Organizational need. Weightings to the Criteria The priorities or weights to the different criteria may be obtained by :

This problem thou Units. Eg. C1 ma comparative term

In order to over c process of norma


2

There are many m

Mutual consultations or opinion polls Pair wise comparison between criteria Establishing a hierarchy of priorities and using AHP ( Analytical Hierarchy Process ) developed by Saaty, which is a well developed computer package in brand name THE EXPERT CHOICE. Criteria for Project Selection Worst-------Best C1 Investment ( in Lakhs of Rs. ) 10 2 C2 Internal Rate of Return( Percentage)10 40 C3 Pay back (Years) 10 2 C4 Risk( V.High, High, Medium, Low, V.low) C5 Future growth (V.Poor, Poor, Medium, good, V.Good) C6 Similarity to existing business (V.Poor, Poor, Medium, Good, V.Good) Note: Each of these criteria is going from certain worst to certain best In different measures !!

Example of a Fighter Aircraft Selection X1 X2 X3 X4 X5 X6 Maximum Speed : Mach Number Ferry Range : Nautical Miles Maximum Pay loads : Lbs Acquisition Cost :Millions of Dollars Reliability : High Low Maneuverability : High Low

The above are, some of the criteria for selection of a Fighter Aircraft

C1
3

Selection of Fig

This example is for NORMALIZ Selection of Fig

X1 Our aim is to Min


4

The first step 2.0 is t A1

TOPSIS Techn Solution :

Step 1: Obtain th Aircraft Selection Intangibles:

X1 Step 2: Obtain No
5

2.0

Aircraft Selection
These weights could be obtained by , subjective opinion polls, Pair level comparison or by AHP methods. If we examine this matrix we can notice the BEST ( high lighted ) and the WORST ( underlined ) for each criteria in each column. Please note that X4 is the Cost criterion, and hence the LEAST value in the Column is the BEST and the HIGHEST value is the WORST. Hence from the weighted decision matrix, we can identify the best and worst options from the four available options for the Aircraft selection. TOPSIS is unique in a sense it looks at what could have been the BEST and What Could have been the WORST. In fact it can measure as to how far the BEST solution from the UTOPIA. ( Utopia is the one which is the IDEAL Solution) We will now proceed to find out how to find the Alternative which is closer to UTOPIA. IDEAL and NEGATIVE IDEAL SOLUTIONS Step 4: Obtain the ideal A* and the Negative Ideal A- solutions from the weighted decision matrix V. A* = (0.1168, 0.0659, 0.0531, 0.0414, 0.1347, 0.2012) A- = (0.0841, 0.0366, 0.0455, 0.0598, 0.0577, 0.1118) Separation Measure Step 5: Compute the separation measures from the Ideal ( Si*) and the negative ideal ( Si - ) solution for all alternatives I =1..m Si* = Sq root ( Sum of separate for J = 1.. N of (Vij Vj*)) Si- = Sq root ( Sum of squares for J = 1..N of (Vij Vj - )) Aircraft Selection Example ( Contd..): Values of Separation Measures :

Once the measure Dominant Value n

We now have the scheme that is ado Similarity to Idea


6

Ideal Solution S1* = 0.0545 S2* = 0.1197 S3* = 0.0580 S4* = 0.1009

Negative Ideal Solution S1- = 0.0983 S2- = 0.0439 S3- = 0.0920 S4- = 0.0458

Relative Closeness to Ideal Solution Step 6: For each alternative determine the relative closeness to the Ideal Solution Ci* , I = 1.. M ) as Ci* = Si - / Si* + Si-)

Aircraft Selec
Relative Closeness Values C1* = 0.643 C2* = 0.268 C3* = 0.613 C4* = 0.312 Note : The closeness rating is a number between 0 and 1, with zero being the WORST possible and 1, the BEST possible solution. From the closeness value numbers for the Aircraft selection Example, we note that the first option gives us the closeness rating of 64.3% to the best possible solution and the second option with only 26.8% is the WORST possible for selection. Rank the Preference Order Step 7: Determine the preference order by arranging the alternatives in the descending order of Ci* , I = 1..m 7

Thus the Ranks for the alternatives in the Fighter Aircraft Selection example using the TOPSIS emerges as : A1, A3, A4, A2.

There is a second called the SAW


NORMALIZATION Step 2: Obtain the Normalized Decision Matrix : R ( r ij , I = 1..m : j= 1..n ) using rij = X ij / X j*, if the jth Criterion is a benefit Criterion and r ij = Xj - / X ij , if the j th Criterion is a Cost Criterion. Patently this is very simple. Let us look at the Decision Matrix once again. In column X1, 2.5 is the best and hence divide everything by 2.5. Use the same procedure for X2, X3, X5 and for X6. But for X4, which is the Cost Criteria Column, 4.5 is the least value, and hence we resort to reciprocal division in that column. That is 4.5 / 5.5, 4.5 / 6.5, 4.5 / 4.5, and 4.5 / 5.5. Subsequently we can obtain the final Scores through the next two steps. Final Scores Step 3: Using the weights for the different criteria obtain the weighted score for each alternative using the Normalized Decision Matrix.

SIM

Step 4: Based on the final Score, rank the alternatives for a decision by the Decision Maker. SAW & TOPSIS ( A Comparison ) 1. Please note that the Rankings obtained by two Multi-Attribute Decision Making techniques need not be identical ( MADM ) 2. For the fighter aircraft selection project SAW gave a rating : A3, A1, A4, A2 and TOPSIS gave a rating: A1, A3, A4, A2 3. The reason for this obviously a) We vary the normalization process for both these schemes

This procedure

Step 1: Obtain t numbers.


8

b) Their pattern of scoring too were different. In the TOPSIS we picked up the BEST solution on the basis of Similarity to Ideal Solution, and in SAW we have taken the weighted average of the score. Summary of Lesson 5 What we learnt in this lesson are: Project selection involves consideration of multiple, often conflicting criteria among different alternatives. ( some times we can evolve a more easier alternative, such as having a cut off price for the tender, with multiple criteria. So out of different vendors who quote for the tender with all the criteria set by us, we can just have one single criteria to evaluate the tender with cut off price only ! ) Project Appraisal leads to evaluations which may be tangible, incommensurate or intangibles. Intangibles are evaluated on a numerical subjective scale. A decision matrix, with numerical evaluations is the starting point for MADM ( Multi Attribute Decision Making ) method. Different methods vary in their normalization schemes and in the manner in which scoring of alternatives is done A fighter aircraft selection example using TOPSIS and SAW is used Next Lesson (6): Project Planning and Scheduling

You might also like