Beruflich Dokumente
Kultur Dokumente
DOI: 10.1243/09544054JEM629
Abstract: Supplier selection has increasingly been regarded as one of the most important
strategies in the globalization era. While many companies require different resources from
the supply market, airline companies require supplies of parts and repair and maintenance
services. The suppliers of these products and services are regarded as critical for the industry.
With the multidimensional nature of the problem, supplier selection involves both tangible
and intangible selection criteria. This paper presents a case study on solving the supplier
selection problem in the airline industry through a decision support system that employs
the analytical hierarchy process (AHP). The system was implemented with the aid of the
commercial software package Expert Choice. The significance of the study is to provide a
new platform for airline companies in selecting suppliers on the basis of the proposed
decision support system, which was tested with a Hong Kong-based airline company during
the course of development. In addition, the model highlights the areas of most concern to
airline companies in supplier selection. Unlike many similar studies, sensitivity analysis has
been performed in order to improve the robustness of the results, and to understand the
importance of different criteria with respect to the selected decision.
Keywords: decision support system, airline industry, case study, supplier selection, analytical
hierarchy process
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Nevertheless, severe intra- and interalliance compe- to wear out, and failures could possibly occur.
tition still exists. Maintenance is then performed to ensure that an
As consumers are quite accustomed to looking for acceptable level of safety is achieved throughout
low fare opportunities, and because the air transport the life of the aircraft and its constituent parts
product is rather difficult to differentiate [6], airline and components. Manufacturing, maintenance, and
companies have been exploiting internet techno- repair errors can negate design safety features and
logies at the customer interface for promotion and introduce hazards that may not be immediately
sales of travel services. In spite of this, it is still quite apparent [11]. Hence, people who are responsible
unlikely that airlines would be able to gain suffi- for these functions play a vital role in company sur-
cient improvement on the revenue side only. As a vival since any notable incident would deteriorate
consequence, more attention should be paid to the the customer impression of air safety. Substantial
control or reduction of operational costs [7], for expenses have been incurred in maintenance, mate-
example, by selecting the best suppliers (which rials, and overhaul by some world-class airlines.
may not necessarily be those offering the cheapest Maintenance deficiencies have been estimated to
products). be involved in approximately 12 per cent of major
Many companies have realized the importance of aircraft accidents and 50 per cent of engine-related
their suppliers to the success of their long-term flight delays and cancellations [12]. This highlights
growth. Hence, they have been striving to develop the importance of suppliers in the airline industry.
and maintain an optimum supplier structure that A number of reports in the literature have pointed
could be beneficial to both sides. The first step in out that one important aspect of supply chain man-
achieving this is supplier-based management, agement is the selection of suppliers [13–15]. In the
which includes supplier evaluation, with subsequent past, many firms have tended to use quantitative
selection of the best subset and reduction in the size factors, such as delivery timeliness, cost, etc., in the
of the base [8]. In other words, supplier evaluation selection of suppliers. Recently, research has shown
and selection are prerequisites for the success of that more companies have adopted channel rela-
any supplier strategy. Given the importance of sup- tionships (i.e. buyer–supplier relationships) in their
pliers and the selection of suppliers in a rather gen- supplier evaluation programmes, or in forming part-
eral perspective, suppliers to the airline companies nerships with suppliers [16]. Wagner [8] advocated
were defined as component or spare parts sup- that the value being generated by a company does
pliers and technical repairers or maintenance service not come solely from within the company or from
providers in the context of the present study. the company’s purchasing or related functions. It
The objective of this study is to design a decision is, to a certain extent, defined by the supply market.
support system (DSS) based on the analytical hier- Therefore, suppliers can significantly contribute to
archy process (AHP) to solve the supplier selection the company’s value. This also highlights the neces-
problem in the airline industry, with the assist- sity and importance of having and maintaining a
ance of the commercial software Expert Choice [9] good supplier-based management system.
to achieve this purpose. AHP has been employed However, relational factors, such as the feeling of
in many applications in DSSs to help management trust, the level and openness of communication,
make decisions [10]. The organization of this paper etc., are likely to be difficult to quantify. Regarding
is as follows. Section 2 reviews related literature. these aspects, Barbarosoglu and Yazgac [15] noted
Section 3 gives a brief introduction of the methodo- that evaluation of intangible factors requires expert
logy of this study. Section 4 presents the DSS, while judgements. Usually, the more a person knows
section 5 is a complementary section to section 4, about the situation, the better and more reliable
discussing related issues by conducting a sensitivity result can be expected from that person. Therefore,
analysis. Sensitivity analysis is important in the supplier selection is certainly a multidimensional
sense that it provides an understanding of the problem involving both tangible and intangible fac-
importance of the criteria, and hence could improve tors, making it more complex to solve [17–19]. In
the robustness of the results. However, it is usually this connection, in order to integrate both tangible
overlooked in the literature. Finally, section 6 con- and intangible factors into a logical and effective fra-
cludes the paper. mework, a DSS based on AHP is proposed in this
study. The system provides a framework to cope
with a multicriteria situation, involving intuitive,
2 LITERATURE REVIEW rational, qualitative, and quantitative aspects, to
solve the supplier selection problem.
The level of safety of an aircraft is initially set by the In fact, a number of authors conducted research
airworthiness standards to which it is designed and into quantifying some intangible factors so that
built. However, during their operating life, parts start a quantitative approach could be employed for
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
decision-making as early as in the 1980s [20]. comparison process since one element in a cluster
Recently, Degraeve et al. [21] developed a mathe- could be compared with other elements in the
matical programming model by using total cost of same cluster more than once. These pairwise com-
ownership for airline selection of a case company. parisons are used to derive ‘local’ priorities of the
The theory behind this is to quantify all the factors elements in a cluster with respect to their parent.
associated with the purchasing process in cost The principle of hierarchical composition or synth-
items. As a matter of fact, not all intangible factors esis is applied to multiply the local priorities of ele-
could be quantified. That is why AHP is still appli- ments in a cluster by the ‘global’ priority of the
cable in some situations and is employed in this parent, producing global priorities throughout the
case study in order to take such unquantifiable hierarchy and then adding the global priorities for
factors into consideration. In fact, ‘quantifying the lowest-level elements (the alternatives).
approaches’ are useful in quantifying some factors As a matter of fact, AHP has been widely applied
for pairwise comparison (to be discussed later) in by a number of researchers and practitioners to
the proposed AHP model so that subjectivity of the solve multicriterion problems, albeit in diverse
model could be further reduced as much as pos- areas [27–33]. For instance, Hafeez et al. [31] used
sible. Nevertheless, it is not our intention to review AHP to determine key capabilities of a firm,
such methodologies in this paper. However, inter- whereas Cheng and Li [25] employed AHP to allo-
ested readers should refer to a relevant review cate resources by prioritizing construction infor-
conducted by De Boer et al. [22] for further mation and to determine the key information. In
reference. particular, in the field of supplier or vendor selec-
AHP, which was developed by Saaty [23], is a tion, AHP has also been suggested and applied by
decision-making process to help people in setting a number of researchers [34, 35]. They suggested
priorities and making the best decisions when that AHP be applied in this field mainly because of
both quantitative and qualitative aspects need to be its inherent capability to take both qualitative and
considered. It allows decision-makers to model a quantitative criteria into consideration in vendor
complex problem in a hierarchical structure, which selection problems.
shows the relationship of the goal of the problem, Some researchers have formulated and used a
assessment areas, objectives (criteria), subobjectives criterion framework to evaluate suppliers and sub-
(subcriteria), and alternatives, like that presented in sequently select the best alternative [15, 18, 34,
Fig. 1. That is, it allows the complex problem to be 36]. On the other hand, others have performed
broken down into smaller parts at different levels. surveys with intended targets, and then developed
Uncertainties and other influencing factors can evaluation criteria or supplier management tools
also be included. Rebstock and Kaula [24] described that are currently used by the target groups of
the established hierarchy as a flexible and dynamic companies [13, 16, 37], rather than using a supplier
model that can be modified if new information or selection model. There is some discussion regard-
insight is acquired. Decision-makers then make ing the criteria determining the process, and the
simple pairwise judgements, typically with redun- number of criteria involved in the AHP. In a study
dancy throughout the hierarchy, to arrive at overall on the supplier selection process using the analytic
priorities for the alternatives. AHP works by first network process (ANP), a generalized form of AHP,
structuring the problem into a form of hierarchy Sarkis and Talluri [18] concluded that the choice
with several levels. Then all the elements in a level and the number of factors to be included in the
are pairwisely compared with respect to the ele- supplier selection process must be cautiously
ments at the level above. This process leads to higher determined, and a large number of factors (40 in
consistency since pairwise comparison requires the that study) and relationships in the model could
decision-maker to think precisely before giving the make decision-makers feel exhausted. Moreover,
answer [25]. they pointed out that reducing the dimensionality
AHP is based on three basic principles: decom- of the problem to an acceptable minimum level
position, comparative judgements, and hierarchical will speed up the decision-making process and
composition or synthesis of priorities [26]. The make the process less complex. In addition, in
decomposition principle, which was mentioned pre- another study that applied AHP to solve vendor
viously, is applied to structure a complex problem selection problems in telecommunications sys-
into a hierarchy of clusters (levels), subclusters, sub- tems, Tam and Rao Tummala [34] noted that the
subclusters, and so on. The principle of comparative presence of too many criteria makes the pairwise
judgements is applied to construct pairwise com- comparison in evaluating vendors a difficult and
parisons of all combinations of the elements in a time-consuming process, which may also lead to
cluster with respect to the parent of that cluster. evaluator assessment bias. In their study, Tam and
Redundancy normally ensues during the pairwise Rao Tummala [34] used 33 criteria. Min [36] also
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
Data and information were collected from various worst. Result analysis was performed with the func-
sources, including international journals, reference tion provided by the software. Sensitivity analyses
books, websites, airline company annual reports, mat- were also performed to investigate the sensitivity of
erials from aviation-related organizations/authorities, the alternatives (i.e. supplier candidates) with
and so on. A reference number was assigned to each respect to the changes in the priorities of the criteria.
material once it had been collected and examined. Expert Choice takes the presumption out of
Analysis was then carried out to form the analytical decision-making based on AHP. It provides a syste-
hierarchical model. matic approach to organizing problems, resulting
An information database was established for the in informed logical decisions. It enables a decision-
purpose of facilitating the analysis process. After maker graphically to portray the relationship between
reviewing each material, the possible criteria to be many facets of a complex problem, and to incorpo-
applied to the airline industry were accommodated rate both quantitative and qualitative information,
in the information database. The information data- including experience and intuition. A hierarchy is a
base accommodated various types of information representation of a complex problem in a multilevel
about each journal, including assigned journal num- structure whose first level is the goal followed succes-
ber, topic of the paper, name of journal, publishing sively by levels of factors, criteria, subcriteria, and
year, criteria being used or possible criteria, and other so on, down to a bottom level of alternatives. In the
important points and their implications. In addition, next two sections, the proposed decision support
the current supplier management framework being system is presented and discussed.
used by the case study company was also included
in the database.
After setting up the database, with the data pre-
4 PROPOSED DECISION SUPPORT SYSTEM
filtered at the examination or reviewing stage, com-
FOR SUPPLIER SELECTION
parisons were initially carried out across the journal
database. This process generated a large number
4.1 AHP model
of criteria. Then, the set of criteria was compared
against the supplier management framework of The model goal, which was placed at the top level,
the case study company, together with other mate- was to select the best supplier for the airline industry
rials, such as publications from the Air Transport (see Fig. 1). The second level was composed of three
Association. With the consideration of the airline main assessment aspects or areas: performance
industry characteristics and its needs, further filtra- assessment, continuous improvement and innova-
tion and selection of the criteria were performed, tion, and company background/business structure.
leading to the formation of a criterion hierarchy in In each area, a number of criteria were identified,
this study. At the same time, a number of questions as discussed in section 3. Performance assessment
were raised during the analysis process. In response includes four criteria: cost, satisfaction of supplier
to this, these questions were sent to the procurement service, quality, and assurance of supply. Under
department of the case study company. In addition, continuous improvement and innovation, a further
the feedback was used for adjusting the final set of four criteria are included in the model: proactive in
criteria. Figure 1 presents the proposed AHP-based measuring supply chain/own suppliers, e-business
model. initiatives, continuous improvement programme/
The next phase was data input and evaluation – TQM/six-sigma programme, etc., and research and
the criterion hierarchy was input to the commercial development (R&D). Under company background/
software Expert Choice [9]. This is a computerized business structure, six criteria are considered:
DSS for analysis, synthesis, and justification of com- organizational culture and strategic issues, compati-
plex decisions and evaluations. In fact, it has been bility (buyer and supplier partnership), perceived
used by different organizations and governments risk, financial issues, technological issues, and
for a wide variety of applications, including alloca- safety, environmental, and education issues. A total
ting resources, value pricing, selecting alternative, of 36 subcriteria were identified. Explanation and
cost–benefit analysis, supplier evaluation, credit justification of each criterion and subcriterion are
analysis, etc. As of 1995, Expert Choice was being listed in Table 1.
used in 57 countries throughout the world, and there For the sake of simplicity, it is assumed that there
were over 1000 journal and other citations about are only five potential suppliers in this study, abbre-
AHP [38]. Pairwise comparisons and/or direct data viated as supp1, supp2, supp3, supp4, and supp 5
input were then made for every element at every in subsequent discussion. They are competing to
level against other elements at that level. All judge- be the best supplier in the assessment of the case
ments were then synthesized, with the alternatives study company. They are real suppliers that
(i.e. supplier candidates) prioritized from best to were drawn from the existing supplier base of the
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Table 1 Explanations and justifications of the criteria and subcriteria in the proposed AHP model
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
Table 1 (Continued)
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Table 1 (Continued)
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
Table 1 (Continued)
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Table 1 (Continued)
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
At the same time, the consistency ratio of each the portion of the goal node priority that the node
judgement was checked to ensure that it is lower possesses (the global priorities of the children of a
than or equal to 10 per cent in order to help the node add up to the global priority of the parent
decision-makers reduce inconsistencies in pairwise node).
comparisons, as suggested by Saaty [23]. It is noted
that, during the judgement process, it is impractical
to pursue zero inconsistent ratios, i.e. perfect consis- 4.3 Synthesis to identify the ‘best’ alternative
tency. One of the local priorities derived is presented After pairwise relative comparisons had been per-
in Fig. 5. formed, the local priorities were synthesized from
Local priority refers to the priority of a node rela- the goal, and the overall priorities were calculated.
tive to its parent node. It represents the percentage The overall priorities of each of the five suppliers
of the parent node priority that is inherited by the are presented in Fig. 6. The distributive mode was
child (the local priorities of the children of a parent selected since the rankings of other suppliers were
node add up to 1). Global priority refers to the prior- of interest to the decision-maker. Supplier 4 received
ity of each node relative to the goal. It represents the highest rating of 0.222. Therefore, it is recognized
as the best supplier. Next came supplier 1, with a
score of 0.207, followed by supplier 3 with a score
Table 2 Pairwise comparison: types and modes of of 0.203, then supplier 2 with a score of 0.189, and
comparison offered in Expert Choice last was supplier 5 with the lowest score of 0.179.
Comment from Expert Choice
manuals
Types of comparison
4.4 Examination and verification of the decision
Importance Appropriate when comparing one
criterion (objective) with another
After the overall priorities had been derived, analysis
Preference Used when comparing the was performed to investigate the sensitivity of the
alternatives ranking of the alternatives to changes in the impor-
Likelihood Used when comparing the
probability of outcomes. It may
tance of the criteria. Sensitivity analysis from the
be used with either criteria or
alternatives
Modes of comparison
Verbal Compare elements (nodes) using Table 3 Priorities of primary assessment objective with
English language terms respect to goal
Graphical Compare elements (nodes) using a
graphical gauge Primary assessment objective Priority
Numerical Compare elements (nodes) using
numbers in a matrix or a Performance assessment 0.531
questionnaire format Continuous improvement and innovation 0.242
Company background and business structure 0.227
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
Table 4 Global priorities of criteria (under each primary Table 5 Global priorities of subcriteria (under each
assessment objective) criterion)
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
Fig. 5 Derived local priorities with respect to parent, performance assessment (note that the sum of
the priorities of cost (A), satisfaction of supplier surface (B), quality (C), and assurance of supply
(D) is equal to 1 since these are local priorities with respect to the parent node, performance
assessment)
Fig. 6 Overall priorities of alternatives in distributive mode (note that the distributive mode is selected
(the distributive icon is pressed))
the alternatives), it identified the impact of changes In Fig. 7, the solid vertical line shows the original
in nodes immediately beneath the selected node on priority of the objective. The gradient (slope) of each
the alternatives. of the alternative lines represents the rate of change
The results of the sensitivity analysis at the goal in priority for the alternatives as the priority of the
level are presented in Figs 7 to 9 (in the form of objective is changed. The representation of the alter-
gradient–sensitivity graphs). They show the sensitiv- natives is labelled as five tabs on the right-hand side.
ity of three primary assessment objectives, i.e. per- The priority of each alternative is the intersection
formance assessment, continuous improvement between the respective lines and the solid vertical
and innovation, and company background and busi- line in Fig. 7. Thus, current ranking is supplier 1 >
ness structure. Figure 10 presents a one-node level supplier 4 > supplier 3 > supplier 2 > supplier 5.
sensitive analysis in a different approach by using a As already stated, the gradient (slope) of each of
performance sensitivity graph. the alternative lines represents the rate of change in
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
priority for the alternatives as the priority of objec- rankings remain unchanged were identified (as
tives is changed. Taking Fig. 7 as an example again, shown within the dotted vertical lines). These are
the priority of supplier 4 (supp4) would decrease the areas where the supplier ratings are not sensitive.
while that of supplier 1 (supp1) would increase if In Fig. 7 it can be seen that, within the first three
the priority of performance assessment were increa- areas (i.e. 1 to 3) of performance assessment, the
sed (the vertical line is moved to the right). For each supplier ratings are not sensitive with respect to sup-
gradient analysis graph, the areas where the supplier plier 4. Hence, when the importance (reflected as the
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
Fig. 10 Performance sensitivity graph of node under objective, continuous improvement and
innovation
weighting) of the performance assessment, which is In Fig. 8, again there are four insensitive areas
now 0.531 as per the position of the solid vertical identified. The areas are bounded by dotted lines. It
line, increases to 0.696 or decreases to 0.457, the is obvious that supplier 4 performs almost the best
current ranking of the suppliers does not change. all the time until the importance of continuous
Regarding the performance assessment, supplier 4 improvement drops below 4.9 per cent, i.e. area 1.
is still the best subject in the sensitivity analysis. By In Fig. 9, only two insensitive areas are identified,
the same token, sensitive analyses for the other two but supplier 4 is insensitive in this case. Figures 7
assessment areas were performed and are shown in to 9 show that a ranking of alternatives is sensitive
Figs 8 and 9. to changes in importance of all the other criteria,
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
except for company background and business struc- including the set of criteria and the weightings,
ture. It is observed that the ranking of suppliers should be used as a reference for constructing a
remains more or less the same throughout the unique AHP model for a particular airline, or adjust-
entire range. The overall utility of the alternatives is ment would be made before directly applying
robust with changes in the importance of company the model. Possible sources of variation could be
background and business structure. identified as follows.
Figure 10 presents a performance sensitivity graph
1. Different sizes of the airlines. Large airlines gene-
of supplier performance on nodes under its parent
rally have greater power in negotiation on the
node, continuous improvement and innovation. Per-
initial price, contractual terms, etc., with the sup-
formance sensitivity analysis shows how well each
pliers. The imbalance in power structure between
alternative performs with respect to each criterion.
suppliers and airlines may affect the supplier’s
The importance of criteria is depicted from the verti-
attitude on such an evaluation model.
cal bars, and values can be read from the left-hand
2. Different operational characteristics. The size of
axis. The performance of each alternative with
the aircraft fleet, or the size (frequency and
respect to each of the criteria is depicted by the
length – long haul or short haul) of the routes it
intersection of a sequence of line segments with
operates, partly influences the amount of replace-
the vertical line at each of the criteria. The overall
ment parts needed and the frequency of mainte-
performance of an alternative is depicted by the
nance required. Moreover, the initial supplier set
intersection of the alternative line segments with
may be different regarding the different aircraft
the ‘overall’ vertical line at the right of the graph.
fleet.
Performance of the alternatives can be read from
3. Different view of top management. The top man-
the right-hand axis. Note that alternative line seg-
agement view is very important in determining
ments between the objectives are drawn for visual
the weightings since the future company direc-
clarity only and do not convey any information.
tion is determined largely by them. For example,
This is different from the gradient-sensitive graph
on the progress of e-business, it may vary from
previously discussed.
one airline to another. Therefore, a completely
As shown in Fig. 10, supplier 4 performs the best
different set of weightings may be applied in its
overall. However, this does not imply that supplier 4
unique model.
performs the best on all four criteria. Instead,
4. Changing environment. Because of the changing
supplier 4 performs the best for only two areas,
external environment, the criteria need to be
namely the ‘continuous improvement program/six-
reviewed progressively, and adjustment of weight-
sigma/TQM programme’ and ‘research and devel-
ings is necessary. For example, environmental
opment’. It ranks the worst in ‘e-business initiatives’
concerns may still largely focus on the operation
and third in ‘proactive management in supply
fleet. However, the trend of more and more
chain/its supplier’. Hence, in spite of supplier 4
governments and organizations realizing the
being the overall best supplier, the company could
importance of environmental sustainability to-
see there is a potential for supplier 4 to improve in
wards economic growth may alert airlines of the
those two particular areas. This may be a piece of
need to adjust the weighting on the criterion
particularly important information to a company
‘supplier environmental management’ so as fully
planning to expand its e-business transactions to
to integrate and gain benefit from the supply chain
speed up and simplify the whole administration
effort.
procedure. Similar analysis could be performed on
the other four suppliers. This is a very important
piece of decision support information to the senior 6 CONCLUSIONS
management of the company.
The purpose of this study has been to develop a DSS
by using AHP to solve the supplier selection problem
5.3 Individual airline characteristics and
in the airline industry. A hierarchical model has been
the changing environment
developed. With the use of the AHP-based decision-
Another aspect of this discussion is a more funda- making software Expert Choice, the result showed
mental issue, that is, individual airline characteris- that the performance assessment objective, which
tics. While it is hoped to apply the proposed AHP comprises subcriteria including cost, satisfaction
model to all airlines, the model may not be able of supplier service, quality, and assurance of supply,
to take account of numerous characteristics of was among the most important areas to be
different airline companies, as well as changing addressed, while the other two objectives, continu-
industry needs. Therefore, the proposed model, ous improvement and innovation and company
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007
background and business structure, were perceived problem. Production and Inventory Mgmt J., 1997, 38
as being of roughly equal importance. (1), 14–21.
One limitation of this study is that the developed 16 Simpson, P. M., Siguaw, J. A., and White, S. C. Measur-
model for finding evaluation criteria is mainly based ing the performance of suppliers: an analysis of evalua-
tion processes. J. Supply Chain Mgmt, 2002, 38(1),
on literature sources or the internet, and on advice
29–41.
given by the case study company. However, it might 17 Weber, C. A., Current, J. R., and Benton, W. C. Vendor
still be possible to use another approach (e.g. a sur- selection criteria and methods. Eur. J. Opl Res., 1991,
vey) to augment the list of criteria used within the 50(1), 2–18.
industry, which is a future line of research of this 18 Sarkis, J. and Talluri, S. A model for strategic sup-
study. No matter how, airline companies are recom- plier selection. J. Supply Chain Mgmt, 2002, 38(1),
mended to make necessary adjustment of the 18–28.
proposed model prior to applying it directly. 19 Ghodsypour, S. H. and O’Brien, C. A decision support
system for supplier selection using an integrated analy-
tic hierarchy process and linear programming. Int. J.
REFERENCES Prod. Economics, 1998, 56–57, 199–212.
20 Primrose, P. L. and Leonard, R. The financial
1 Condom, P. Airline industry performance: past, pre- evaluation and economic application of advanced
sent and future. In International air transport: the manufacturing technology. Proc. Instn Mech. Engrs,
challenges ahead, 1993, pp. 21–44 (Organization for Part B: Management and Engineering Manufacture,
Economic Co-Operation and Development, Paris). 1986, 200(1), 27–31.
2 Boeing Company Current Market Outlook 2004. 21 Degraeve, Z., Labro, E., and Roodhooft, F. Total cost
Available at http://www.boeing.com/commercial/ of ownership purchasing of a service: the case of airline
cmo/pdf/CMO2004.pdf selection at Alcatel Bell. Eur. J. Opl Res., 2004, 156(1),
3 Kangis, P. and O’Reilly, D. Strategic responses to com- 23–40.
petitive pressures: European Air Transport. J. Strategic 22 De Boer, L., Labro, E., and Morlacchi, P. A review of
Change, 1998, 7(3), 167–182. methods supporting supplier selection. Eur. J. Purchas-
4 Taneja, N. K. Driving airline business strategies through ing and Supply Mgmt, 2001, 7(2), 75–89.
emerging technology, 2002 (Ashgate Publishing Limited, 23 Saaty, T. L. The analytic hierarchy process, 1980
Aldershot). (McGraw-Hill, New York).
5 Bennett, M. M. Strategic alliances in the world airline 24 Rebstock, S. E. and Kaula, R. The effectiveness of an
industry. Prog. in Tourism and Hospitality Res., 1997, analytical hierarchy process in group decision making:
3(3), 213–223. a case study. Int. J. Computer Applic. in Technol., 1996,
6 O’Connor, W. E. An introduction to airline economics, 9(2/3), 95–105.
2001 (Praeger, Westport, Connecticut). 25 Cheng, E. W. L. and Li, H. Information priority-setting
7 Seristo, H. The executive view on the cost problem for better resource allocation using analytic hierarchy
of European airlines. Eur. Business Rev., 1996, 96(4), process (AHP). Inf. Mgmt and Computer Security,
14–17. 2001, 9(2), 61–70.
8 Wagner, S. M. A strategic approach to professional 26 Saaty, T. L. Fundamentals of decision making and
supplier management. Natn. Productivity Rev., 2000, priority theory with the analytic hierarchy process,
19(3), 21–28. 1994 (RWS Publications, Pittsburgh).
9 EC Pro for Windows decision support software tutorial, 27 Ngai, E. W. T. Selection of web sites for online adver-
Version 9.0, 1986 (Expert Choice Inc., Virginia, USA). tising using the AHP. Inf. and Mgmt, 2003, 40(4),
10 Changchien, S. W. and Lin, M. C. Design and imple- 233–242.
mentation of a case-based reasoning system for mar- 28 Lai, V. S., Wong, B. K., and Cheung, W. Group decision
keting plans. Expert Syst. with Applic., 2005, 28(1), making in multiple criteria environment: a case using
43–53. the AHP in software selection. Eur. J. Opl Res., 2002,
11 Wells, A. T. Commercial aviation safety, 2000 137(1), 134–144.
(McGraw-Hill, New York). 29 Byun, D.-H. The AHP approach for selecting an
12 Marx, D. A. and Graeber, R. C. Human error in aircraft automobile purchase model. Inf. and Mgmt, 2001,
maintenance. In Aviation psychology in practice (Eds 38(5), 289–297.
N. Johnston, N. McDonald, and R. Fuller), 1994, 30 Duke, J. M. and Aull-Hyde, R. Identifying public pre-
pp. 87–104 (Avebury Technical, Aldershot). ferences for land preservation using the analytical hier-
13 Park, D. and Krishnan, H. A. Understanding supplier archy process. Ecological Economics, 2002, 42(1–2),
selection practices: differences between US and Korean 131–145.
executives. Thunderbird Int. Business Rev., 2001, 43(2), 31 Hafeez, K., Zhang, Y. B., and Malak, N. Determining
243–255. key capabilities of a firm using analytical hierarchy
14 Mccollum, B. D. How changing purchasing can change process. Int. J. Prod. Economics, 2002, 76(1), 39–51.
your business. Production and Inventory Mgmt J., 2001, 32 Muralidhar, K., Santhanam, R., and Wilson, R. L.
42(2), 57–60. Using the analytic hierarchy process for information
15 Barbarosoglu, G. and Yazgac, T. An application of the system project selection. Information and Mgmt,
analytic hierarchy process to the supplier selection 1990, 18(2), 87–95.
JEM629 Ó IMechE 2007 Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture
33 Ta, H. P. and Har, K. Y. A study of bank selection strategies, and case studies, 2003 (McGraw-Hill/Irwin,
decisions in Singapore using analytical hierarchy Boston, Massachusetts).
process. Int. J. Bank Marketing, 2000, 18(4), 170–180. 42 Radosevich, L. Bringing home tomorrow’s bright ideas.
34 Tam, M. C. Y. and Rao Tummala, V. M. An application InfoWorld, 1998, 20(43), 71–72.
of AHP in vendor selection of a telecommunications 43 McCormack, K. P., Johnson, W. C., and Walker, W. T.
system. Omega – The Int. J. Mgmt Sci., 2001, 29(2), Supply chain networks and business process
171–182. orientation: advanced strategies and best practices,
35 Nydick, R. L. and Hill, R. P. Using the analytic hierar- 2002 (St Lucie Press, Boca Raton, Florida).
chy process to structure the supplier selection proce- 44 Anderson, E. and Weitz, B. The use of pledges to build
dure. Int. J. Purchasing and Mater. Mgmt, 1992, and sustain commitment in distribution channels.
28(2), 31–36. J. Marketing Res., 1992, 29, 18–34.
36 Min, H. International supplier selection: a multi- 45 Robbins, S. P. and Coulter, M. Management,
attribute utility approach. Int. J. Phys. Distribution 2002 (Prentice-Hall, Upper Saddle River, New
and Logistics Mgmt, 1994, 24(5), 24–33. Jersey).
37 Tan, K. C. Supply chain management: practices, 46 Kamata, E. S. Influence of psychological factors on
concerns, and performance issues. J. Supply Chain product development: lessons from aerospace and
Mgmt, 2002, 38(1), 42–53. other industries, 2002 (Kluwer Academic Publishers,
38 Forman, E. H. and Selly, M. A. Decision by objectives: Dordrecht).
how to convince others that you are right, 2001 (World 47 Chen, C. C., Chen, X.-P., and Meindl, J. R. How can
Scientific, Singapore). co-operation be fostered? The cultural effects of
39 McIvor, R., O’Reilly, D., and Ponsonby, S. The impact individualism-collectivism. Acad. Mgmt Rev., 1998,
of Internet technologies on the airline industry: current 23(2), 285–304.
strategies and future developments. Strategic Change, 48 Hobbs, A. and Williamson, A. Unsafe acts and unsafe
2003, 12(1), 31–47. outcomes in aircraft maintenance. Ergonomics, 2002,
40 Djamaludin, I., Murthy, D. N. P., and Kim, C. S. 45(12), 866–882.
Warranty and preventive maintenance. Int. J. Reliabil- 49 Huselid, M. A. The impact of human resource manage-
ity, Qual. and Saf. Engng, 2001, 8(2), 89–107. ment practices on turnover, productivity, and corpo-
41 Simchi-Levi, D., Kaminsky, P., and Simchi-Levi, S. rate financial performance. Acad. Mgmt J., 1995,
Designing and managing the supply chain: concepts, 38(3), 635–672.
Proc. IMechE Vol. 221 Part B: J. Engineering Manufacture JEM629 Ó IMechE 2007