See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/259196876

Decision-Making with Multiple Criteria Using AHP and MAUT: An Industrial

Application

Article · January 2013

CITATIONS READS

17 2,487

7 authors, including:

Ana Paula Camilo Pereira Estéfano Vizconde Veraszto

Universidade Estadual de Mato Grosso do Sul Universidade Federal de São Carlos
40 PUBLICATIONS   199 CITATIONS    199 PUBLICATIONS   440 CITATIONS   

SEE PROFILE SEE PROFILE

Fernando Augusto Silva Marins Messias Borges Silva

São Paulo State University University of São Paulo USP and São Paulo State University UNESP
194 PUBLICATIONS   1,278 CITATIONS    281 PUBLICATIONS   1,537 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

INDUSTRY 4.0 IMPACTS ON WORKERS: CASE STUDY OF THE BRAZILIAN, BRITISH, CANADIAN and CHINESE AUTOMOTIVE INDUSTRY View project

Strategic Management View project

All content following this page was uploaded by Messias Borges Silva on 16 May 2014.

The user has requested enhancement of the downloaded file.

European International Journal of Science and Technology Vol. 2 No. 9 November, 2013

Decision-Making with Multiple Criteria Using AHP and MAUT:

An Industrial Application

Leandro Valim de Freitas

Petróleo Brasileiro SA, PETROBRAS, Brazil
São Paulo State University (UNESP), Brazil

Ana Paula Barbosa Rodrigues de Freitas

São Paulo State University (UNESP), Brazil

Estéfano Vizconde Veraszto

Federal São Carlos University (UFSCAR), Brazil

Fernando Augusto Silva Marins

São Paulo State University (UNESP), Brazil

Messias Borges Silva

São Paulo State University (UNESP), Brazil
São Paulo University (USP), Brazil

Abstract
This work presents the modeling of decision making with multiple criteria for the selection of suppliers in a
metallurgical company located in the state of Sao Paulo. It was needed to systematize procedures for
purchasing raw materials, taking price, quality, and delivery factors into account from three potential
suppliers. MAUT and AHP methods were compared. Summarizing the results using the MAUT method, a
sensitivity analysis was performed for the criterion Quality. With the AHP Method, it was possible to assess
that the judgments shown are consistent. In both methods, the vector of decision was calculated and the
results corroborate thevsuggestionvof supplier B.

Key words: Decision-Making, AHP, MAUT

1. Introduction
Decision making with multiple criteria (MCDM, Multiple-Criteria Decision-Making) has been successfully
used in complex business problems, such as supplier selection, which is a theme that stands out because of
its potential impact on the entire supply chain of an organization [1].
There are several methods for decision making, such as Analytic Hierarchy Process), ELECTRE
(Elimination et Choix Traduisant la Réalité), MACBETH (Measuring Attractiveness by a Categorical Based
Evaluation Technique) and MAUT (Multiple Attribute Utility Theory) [2] and are used around the world,
such as in governments, businesses, industries, health and educational institutions.
The objective of this work is to compare different techniques of multicriteria decision aid, specifically, the
AHP and the MAUT in mathematical modeling for raw materials supplier selection in a metallurgical plant
located in Brazil.
93
European International Journal of Science and Technology ISSN: 2304-9693 www.eijst.org.uk

2. Multiple-Criteria Decision-Making

2.1 MAUT (Multiple Attribute Utility Theory)

MAUT uses a function called Utility, which aims to replace the value associated with a criterion for a level
of satisfaction to the decision maker [3].
In this method, the assigning values calculated directly. At the highest level of satisfaction for a given
criterion, it is assigned a value of 1 (Utility equals 1) and, at the lowest one, the value zero is assigned
(Utility equals 0). Intermediate values may be calculated by interpolation.
The relative weight of the criteria should be normalized and the decision vector, D, or the expected utility, is
obtained by multiplying the decision matrix, M, with weight vector criteria, w, according to Equation 1 [4].

D = Mw (1)

Synthesizing the results with MAUT, it is used a Sensitivity Analysis to assess the weight impact of a
certain criterion on the decision. Figure 1 shows the expected utility varies with the weight variation of a
given criterion. For example, alternative C is chosen over alternatives A and B only if the weight of the
criterion in question is greater than 0.6.

Figure 1. Sensitivity Analysis exemplifying chart

2.2 AHP (Analytic Hierarchy Process)

The AHP Method was developed by Thomas A. Saaty in the 70s at University of Pennsylvania [6].
The AHP decomposes their decision problem into a hierarchy of independent sub-problems at different
levels, as shown in Figure 2.

Figure 2. Hierarchical Structure Decision [5]

94
European International Journal of Science and Technology Vol. 2 No. 9 November, 2013

The criteria and alternatives are compared in pairs through comparisons matrices using the Fundamental
Scale of Saaty, according to Table 1:

Table 1. Fundamental Saaty scale [3]

1 Equal Importance
2 Weak or slight
3 Moderate importance
4 Moderate plus
5 Strong importance
6 Strong plus
7 Very strong or demonstrated importance
8 Very, very strong
9 Extreme importance

The importance of the criteria is obtained by Linear Algebra, as in Equation 2 [7] where A represents the
matrix of judgments, w, is the eigenvector, and λmax, is the maximum eigenvalue.

Aw = λmáxw (2)

Among other ways, the vector performance can be obtained by geometric mean.
From the normalized eigenvector of each matrix comparisons, it is obtained the vector of performance or the
priorities of alternatives for each criterion in question.
In the AHP, it is possible to know the degree of consistency between judgments, and the CR (Consistency
Ratio), according to Equation 3 [8]:

CR = (λmax – n) / (n –1) RI (3)

Equation 3 considers a random error, RI (Random Consistency Index), associated with the order n of the
matrix, as shown in Table 2.

Table 2. Random Index [6]

N 3 4 5 6 7 8 9 10
RI 0.52 0.89 1.11 1.25 1.35 1.40 1.45 1.49

3. Application
At this stage, it is reported a case study, aiming to provide higher subsidies for decision making in the choice
of raw material suppliers for company´s board of directors. The entire process of analysis of alternatives has
been based on an economic viability analysis so far, and on the final manufactured products have already
met the desired specifications or not. However, in order to obtain a more competitive performance, the
company should insert a systematic way in the process of buying raw material, and focus on two other
factors besides the price: quality and delivery time.
The problem structure is presented in Table 3:

95
European International Journal of Science and Technology ISSN: 2304-9693 www.eijst.org.uk

Table 3. Problem structure

Supplier (Price) (Quality) (Delivery time)

Criterion C3
Criterion C1 Criterion C2
A $6500 Good Median
B $ 9000 Acceptable Very good
C $ 7500 Very good Bad

The data collection was conducted through questionnaires.

4. Results and Discussion

4.1 Summary results with MAUT

The rating used in the questionnaire, was from 0 (worst) to 4 (best) for a direct assignment of values. Table 4
shows a summary of rhe ratings.

Table 4. Direct assignment of values

Alternative C1 C2 C3
A 4 2 2
B 2 4 4
C 3 3 1

It was adopted a strictly linear preference with neutrality in relation to a risk in the conversion process, and
a definition of intermediate values of utility was presented, as shown in Table 5. The process was conducted
for each criterion:

Table 5. Utility Values

Alternative C1 C2 C3
A 1 0 0.333
B 0 1 1
C 0.5 0.5 0

The criteria vector weight was normalized as shown in Table 6.

Table 6. Vector weight of criteria

Criterion Relative Normalized

Weight Weight
C1 1 1 / 4 = 0.25
C2 2 1 / 2 = 0.5
C3 1 1 / 4 = 0.25
SUM = 4
96
European International Journal of Science and Technology Vol. 2 No. 9 November, 2013

The calculation of the decision vector was performed according to Equation 1, which indicates that vendor B
should be chosen, as in the following matrix notation:

. =

The sensitivity analysis in relation to the quality criterion (C2) revealed that only a normalized weight with
less than 14%, would cause the decider to change the selection from company B´s vendor to vendor A, as
shown in Figure 3.

Figure 3. Sensitivity Analysis – Quality Criterion

4.2 Summary results with AHP

Initially, the criteria are assessed in pairs, according to Saaty´s scale. A summary of the results is shown in
the matrix below:

C1
C1 C2 C3
C2
C3

And next, the geometric mean is performed in the lines, according to equation 3:

After that, vector w of the criteria is obtained by normalizing the geometric mean:

97
European International Journal of Science and Technology ISSN: 2304-9693 www.eijst.org.uk

Then, the alternatives are also compared in pairs for each criterion. With regard to criterion Price (C1), no
judgment in pairs is required, because the monetary amount is already known, however, it is an indirect
attribute, i.e., the lower it is, the better it becomes, so it is necessary to harmonize the results prior to
normalizing them. Furthermore, for criterion C1, the geometric mean calculation is not applied.
The harmonization process is represented by the vector below:

The vector is normalized as:

The judgments matrix to criterion C2 is shown :

This is subject to the calculation of geometric mean and normalization:

Regarding the Delivery time criterions (C3), following the results of judgment matrix, geometric mean, and
normalized weight are follows:

98
European International Journal of Science and Technology Vol. 2 No. 9 November, 2013

The decision vector can also be obtained by Equation 1, and it corroborates the decision suggested by the
previous method.

. =

Finally, the λmáx and CR are obtained by equations 2 and 4:

. 3.006 and CR = 0.0053

5. Conclusions
Decision making with multiple objectives and multiple criteria is a common example in business. In this
study, it was demonstrated an application of MAUT and AHP methods, such as decision support, in an
action research on supplier selection in a metallurgical industry.
In both methods, the vectors of decision pointed at the same answer, the choice of supplier B.
The sensitivity analysis was performed by synthesis with MAUT, and the results showed that supplier B
should be chosen for all weights up to 14%, compared to quality criterion (C2) where the point of
intersection between the straight lines represented alternatives A and B. Just below this point, it indicates
that alternative A should be chosen, which does not reflect the reality of the company because it is
concerned with the perceived quality.
This study demonstrated a comprehensive solution of the AHP, although software can solve the matrix
calculations, and it can be applied more easily, too.
The maximum eigenvalue (λmáx = 3,006) was quite close to the calculated order of the square matrix of
decision (n = 3), indicating little dispersion, which reflects the value of the consistency ratio (CR = 0.53%).

6. Acknowledgements
The authors would like to thank Fundação para o Desenvolvimento Científico e Tecnológico (FDCT) for
translation support.

99
 European International Journal of Science and Technology ISSN: 2304-9693 www.eijst.org.uk

7. References

[1] Kaes, I., Azeem, A. (2009). Demand Forecasting and Supplier Selection for Incoming Material in RMG
Industry: A Case Study. International Journal of Business and Management, 4, 149-159.

[2] Woitowicz, B. B. C. (2009). A Multicriteria Approach for Selecting Business Intelligence Tools.
Dissertation (Master of Administratin), IBMEC.

[3] Shimizu, T. & Jansen, L. K. C., Jansen J. U. (2004). An investment analysis considering intangibles
factors, XXIV National Meeting of Production Engineering, ENEGEP.

[4] Salomon, V. A. P. (2010). Contributions to Validation in Decision Making with multiple criteria. Thesis,
São Paulo State University, UNESP.

[5] Saaty, T. L. (2001). Models, methods, concepts & applications of the Analytic Hierarchy Process.
Kluwer Academic Publishers, Inc., Dordrechet. 345p.

[6] Saaty, T. L. (1980). The Analytic Hierarchy Process. McGraw-Hill, Inc., New York. 287p.

[7] Aguiar, D. C., Souza, H. J. C., Salomon, V. A. (2010). AHP Application to Evaluate Scoring Criteria for
Failure Mode and Effect Analysis (FMEA). International Journal of Analytic Hierarchy Process, 2, 3-13.

[8] Garber, M. F., Szajnbok, M. (2005). Application of Utility Theory in Decision Shipbuilding, XIX
Congresso Panamericano de Ingeniería Naval, Transporte Marítimo e Ingeniería Portuaria.

100

View publication stats