an innovative integration of fuzzy logic and systems dynamics

Computers & Industrial Engineering 88 (2015) 1–12

Contents lists available at ScienceDirect

Computers & Industrial Engineering

journal homepage: www.elsevier .com/ locate/caie

An innovative integration of fuzzy-logic and systems dynamicsin sustainable supplier selection: A case on manufacturing industry

http://dx.doi.org/10.1016/j.cie.2015.06.0190360-8352/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author.E-mail address: [email protected] (I.J. Orji).

Ifeyinwa Juliet Orji ⇑, Sun WeiSchool of Mechanical Engineering, Dalian University of Technology, 116024, China

a r t i c l e i n f o

Article history:Received 13 July 2014Received in revised form 29 April 2015Accepted 12 June 2015Available online 2 July 2015

Keywords:Fuzzy logicSupplier selectionSustainabilitySystems dynamics

a b s t r a c t

Globally, supply chains compete in a complex and rapidly changing environment. Hence, sustainablesupplier selection has become a decisive variable in the firm’s financial success. This requires reliabletools and techniques to select the best sustainable supplier and enhance understanding about howsupplier behavior evolves with time. System dynamics (SD) is an approach to investigate the dynamicbehavior in which the system status alterations correspond to the system variable changes. Fuzzy logicusually solves the challenges of imprecise data and ambiguous human judgment. Thus, this work pre-sents a novel modeling approach of integrating information on supplier behavior in fuzzy environmentwith system dynamics simulation modeling technique which results in a more reliable and responsibledecision support system. Supplier behavior with respect to relevant sustainability criteria in the past,current and future time horizons were sourced through expert interviews and simulated in Vensim toselect the best possible sustainable supplier. Simulation results show that an increase in the rate ofinvestment in sustainability by the different suppliers causes an exponential increase in total sustainabil-ity performance of the suppliers. Also, the growth rate of the total performance of suppliers outruns theirrate of investment in sustainability after about 12 months. A dynamic multi-criteria decision makingmodel was presented to compare results from the systems dynamics model.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Today, industries consider how to manage supply chain opera-tions more efficiently to improve organizational performance.Sustainable supplier selection is a very critical agent to ensuringthe profitability and survival of a company. However, a major prob-lem while implementing sustainable supplier selection is how toinsure that suppliers maintain their status for a long period.

The success of green manufacturing lies hugely on selectingsustainable suppliers. An effective supplier selection model is acritical success factor for supply chains in a complex environmentby providing fast response to the business system changes in pre-dominant dynamic environment. A reliable decision making pro-cess requires understanding of a complex situation of thebusiness. Thus, static mathematical modeling techniques in opera-tions research might not be deemed reliable given their inability tointegrate all the variables of a real situation into the decision sup-port models.

There exist in the literature many approaches to the topic ofsustainable supplier selection most of which are based on

multi-criteria decision making models (Amindoust, Ahmed,Saghafinia, & Bahreininejad, 2012; Awasthi, Chauhan, & Goyal,2010; Bottani & Rizzi, 2008; Büyüközkan & Çifçi, 2011;Campanella, Pereira, Ribeiro, & Varela, 2012; Campanella &Ribeiro, 2011; Chang, Chang, & Wu, 2011; Chen, 2009; Javad,Rita, & Leonilde, 2014; Mani, Agrawal, & Sharma, 2014; Zhang,Hamid, Bakar, & Thoo, 2014). The basic assumption in applyingdecision making models is that both criteria and alternatives arefixed a priori and that decision occurs only once i.e., does notinvolve spatial or temporal considerations. This assumptionundoubtedly limits the validity of the result, specifically whenthe values change over time and the decision matrix is not fixedor static as in sustainable supplier selection problems. In addi-tion, the multi-criteria decision-making model conception of thesupplier selection problem focuses on the cause and effect rela-tionship between the system components individually; it is thusnot regarded as a broad model. Multi-criteria decision makingmodels usually does not provide a complete understanding of thecomplex nature of the supplier selection problem with respect toeconomic, social and environmental factors. Thus, multi-criteriadecision making models cannot reliably provide informationon insuring suppliers maintain their status for a long period oftime.

http://crossmark.crossref.org/dialog/?doi=10.1016/j.cie.2015.06.019&domain=pdf

http://dx.doi.org/10.1016/j.cie.2015.06.019

mailto:[email protected]

http://dx.doi.org/10.1016/j.cie.2015.06.019

http://www.sciencedirect.com/science/journal/03608352

http://www.elsevier.com/locate/caie

2 I.J. Orji, S. Wei / Computers & Industrial Engineering 88 (2015) 1–12

In efforts towards ensuring more effective decision making,research efforts have involved applying soft operation researchmodeling techniques like; strengths weaknesses opportunitiesand threats (SWOT) analysis, decision tree and system dynamics(SD). Techniques that are predominantly rational, interpretative,structure and qualitative are employed by soft operations researchmodels which usually interpret, define, and explore various per-spectives of the problem (Heyer, 2004).

System dynamics is one of the promising soft operationresearch techniques. It was developed from the research carriedout by Jay W. Forrester at the Massachusetts Institute ofTechnology. Jay W. Forrester defines Industrial Dynamics as involv-ing the study of the information feedback characteristics of indus-trial activity to show how organizational structure, amplification(in policies), and time delays (in decision and actions) interact toinfluence the success of the enterprise (Forrester, 1971). The gen-eral belief that systems dynamics modeling is more suitable formodeling at the strategic level was countered by the survey ofTako and Robinson (2012) on the journal articles identified in theperiod (1996–2006) which shows that systems dynamics modelinghas been applied more in the operational level. Systems dynamicsmodeling allows the researcher to analyze complex systems from adynamic viewpoint, rather than from a static perspective. The twomain reasons for System Dynamics popularity are the complex nat-ure of the problem and the qualitative factors such as humanbeings evolvement in those processes (Khatie, Bulgak, & Segovia,2010). The systems dynamics approach considers system as awhole by covering all of the interactions among the componentsof the system. It is a broad approach which incorporates all the ele-ments of a system and thus considered a reliable decision makingapproach. The system dynamics approach has the ability to effec-tively model the feedback and feed forward information in a com-plex dynamic system. The expected outcomes of system dynamicsmodeling are not necessarily quantitative point predictions forparticular variable, but rather a measure of the pattern of dynamicbehavior of the system, given the variables and conditions in themodel (Wareef, 2013). The systems dynamics model incorporatesspatial or temporal considerations and assumes that criteria andalternatives are not fixed. Thus, the validity of the system dynam-ics model results is increased, specifically in sustainable supplierselection where the supplier performance values change over timeand the decision matrix is not static. Therefore, systems dynamicscan be applied in the sustainable supplier selection problem toinsure suppliers can maintain their status for a long period of time.

A good decision-making model needs to tolerate vagueness orambiguity because fuzziness and vagueness are common charac-teristics in many decision-making problems (Lee, Chen, & Chang,2008). In proffering solution to many real world problems (likesupplier selection) that involve some degree of imprecision andambiguity, fuzzy logic is deemed essential (Bayrak, Celebi, &Taskin, 2007; Bevilacqua & Petroni, 2002; Kahraman, Cebeci, &Ulukan, 2003; Ordoobadi, 2009). Fuzzy theory is most preferredto solve the problems of imprecise data and ambiguous humanjudgments in supplier selection (Chang et al., 2011).

In this work, a novel modeling approach for integrating infor-mation on supplier behavior in fuzzy environment with systemdynamics simulation modeling technique was developed to pro-vide insight into how supplier behavior evolves with time. The pro-posed approach represents the total sustainability performance ofsuppliers in the past, present and future period in a green manufac-turing environment. Simulation results show that an increase inthe rate of investment in sustainability by the different supplierscauses an exponential increase in total sustainability performanceof the suppliers. The systems dynamics modeling approach pre-sented in this study can be applied to any green manufacturingenvironment regardless of the number of alternatives and relevant

sustainability criteria. At the development of the systems dynamicmodel, time and cost resources could be demanding, but onceinstalled, use of the model becomes less demanding. The remain-ing parts of this paper will discuss the novel approach which iscapable of: (a) Estimating supplier behavior with respect to sus-tainability criteria in the past, present and future period. (b)Providing insight into how supplier behavior evolves with time.

It is believed that this work can support the selection of sustain-able suppliers and insuring suppliers maintains their status for along period of time.

2. Literature review

Although a rich supplier selection literature exists, there hasbeen relatively little research that investigates how to insure sus-tainable suppliers maintain their status for a long period of time.Most past works employed multi-criteria decision making modelsin solving supplier selection problems. Bottani and Rizzi (2008)integrated fuzzy with cluster analysis and multi-criteria decisionmaking model (MCDM) to solve the supplier selection problem.Awasthi et al. (2010) in their work integrated fuzzy with TOPSISto evaluate environmental performances of suppliers. Wu and Liu(2011) proposed a supplier selection application based on twomethods: VIKOR algorithm and fuzzy TOPSIS with vague setsmethods. Khamseh and Mahmoodi (2014) presented hybrid modelfor green supplier selection based on fuzzy TOPSIS-TODIM employ-ing fuzzy time functions. Aghajani and Ahmadpour (2011) pro-posed fuzzy-TOPSIS for ranking of suppliers in automobilecompanies in Iran. Wang, Cheng, and Huang (2009) presented afuzzy hierarchical TOPSIS for supplier selection which is capableof evaluating uncertainty and choosing the best supplier.Wittstruck and Teuteberg (2011) presented an integrated modelbased on fuzzy-AHP-TOPSIS for recycling partner selection thataccounts for sustainability factors. Azadnia, Saman, and Wong(2015) developed a mathematical programming model for sustain-able supplier selection and order-lot sizing. Büyüközkan and Çifçi(2011) presented a novel model based on fuzzy analytic networkprocess within multi-person decision-making environment undervague preference relations. Their model is able to make effectiveevaluations using available preference information and maintainconsistency level of evaluations. Verdecho, Alfaro-Siaz, and Rodríguez-Rodríguez (2010) proposed a performance managementmodel based on ANP for supplier selection in automotive industryin Spain.

Other decision approaches has been applied to supplier selec-tion problem. Jauhar, Pant, and Abraham (2014) presented a novelapproach for sustainable supplier selection based on differentialevolution to select the efficient sustainable suppliers and providethe maximum fulfillment for the sustainable criteria determinedin a pulp and paper industry. Foerstl, Reuter, Hartmann, andBlome (2010) hinged on the dynamic capabilities view (DCV) topropose that management capabilities of sustainable suppliersare critical agents able to give competitive advantage. However,their approaches do not provide information on whether supplierscan maintain their status for a long period of time.

Several real world examples have proven the interdisciplinarynature and capability of systems dynamics modeling in solving realworld complex problems. Systems dynamics was applied in ana-lyzing the behavior of manufacturing in supply chain(Vashiranwongpinyo, 2010). Systems dynamics simulation wasutilized to analyze the behavior of a generic short life cycle supplychain (Briano, Caballini, Giribone, & Revetria, 2010). The systemsdynamics approach has been widely used to conduct policy exper-iments by many researches and policy makers for over 30 years(Trappey, Trappey, Hsiao, Ou, & Chang, 2012). System dynamics

I.J. Orji, S. Wei / Computers & Industrial Engineering 88 (2015) 1–12 3

models are also often used to address environmental problems andsustainability issues. For instance, wildlife population dynamics,air polluting, and vehicle emissions has been studied (Ford,1999). The global perspectives of environmental sustainabilityissues were contemplated with a broader scope (Forrester, 1971;Meadows, Randers, & Meadows, 1993). The SD approach has beenutilized to investigate the effects of increasing human populationon the earth and natural resources (Meadows, Randers, &Meadows, 2004; Randers, 2000). Several other studies that utilizedsystems dynamics modeling approach includes the issues relatedto regional sustainable development, environmental management,water resource planning, urban planning, and ecological modeling(Onat, Egilmez, & Tatari, 2014).

Vensim software is widely adopted in developing systemdynamics models for simulation applications which provide auser-friendly interface. In addition, it offers a flexible way todynamically map and provide information on how complex sys-tems and inputs really work by building a variety of simulationmodels. Vensim has been applied in electronic commerce riskmechanism research (Qiang, Hui, & Xiao-dong, 2013), environmen-tal modeling (Elsawah et al., 2012), single-stage inventorysystem (Belhajali and Hachicha, 2013), climate system (Akhtar,

Fig. 1. A novel approach for sus

Simonovic, Wibe, MacGee, & Davies, 2011), learning effectivenessevaluation (Lan, Lan, Chen, Chen, & Lin, 2013), andland-use/transport interactions (Haller, Emberger, & Mayerthaler,2008). In this study, the software package of Vensim was employedto the model building of sustainable supplier behavior with respectto sustainability factors in different time horizons. Till date to thebest our knowledge, not much attention has been given to theresearch of insuring suppliers maintain their status for a long per-iod while implementing sustainable supplier selection. Thus, thiswork pioneers the application of integrated fuzzy logic and sys-tems dynamics in the study of sustainable supplier selection.

3. Methodology

The detailed presentation of the novel modeling approach pro-posed for the sustainable supplier selection problem is shown inFig. 1.

Within the development of the proposed model for sustainablesupplier selection, sustainability factors relevant to case study, lin-guistic scale and alternatives were progressively defined.

tainable supplier selection.


3.1. Case study and problem specifications

A gear manufacturing company in China was used as the samplepopulation from which the purchasing unit was selected. Theinformation used for the study was gathered from archival recordsand interviews with personnel. The personnel considered in thisstudy were 17 experts/purchasing managers. The gear manufactur-ing company under study produces 2 MW wind power gear boxand sources for gear wheel shaft as its main raw material. For thisstudy, 4 suppliers are considered as alternatives for providing thegear wheel shaft.

The first step in developing a systems dynamics model is todefine the causal loop diagram. Causal loop diagrams are usefulfor identifying the feedback loops involved in the process and alsodiagramming the feedback structure of systems. Fig. 2 shows acausal loop diagram for the sustainable supplier selection problemfocusing on green design and information disclosure criteria.

The causal loop diagram presented shows the variables for foursuppliers with respect to two sustainability criteria namely greendesign and information disclosure. The similar structure can bereplicated for other relevant sustainability criteria. According tothe developed causal loop diagram, the process consists of feed-back loops namely final performance of suppliers with respect to

Fig. 2. Causal loop diagram for su

sustainability criteria, sustainability pool rate adjustment and esti-mation of supplier performance with respect to specific sustain-ability criteria. This diagram will serve as a basis for developingstock and flow model for sustainable supplier monitoring andselection that will be discussed in Section 4.

The process of estimating supplier performance with respect tospecific sustainability criteria starts with the total budget/invest-ment of a particular supplier. The supplier performance withrespect to sustainability criteria specific time periods is estimatedby experts using fuzzy questionnaires. The aggregation of all theperformances of a particular supplier with respect to the sustain-ability criteria gives the final sustainability performance of thesupplier. The best sustainable supplier can be selected by employ-ing a simple average method to determine the highest final sus-tainability performance of the suppliers in the past, current andfuture time horizons.

The pool rate of sustainability criteria at the supplier level canbe defined as the sustainability performance generation rate foreach supplier with respect to sustainability criteria. The ratedefines the total sustainability performance incurred by supplierfor a certain time horizon. The supplier level rate can be calculatedfrom the total supplier performance and the total percentage ofsustainability criteria driver in the past, current and future time

stainable supplier selection.

Table 2Supplier performance with respect to sustainability sub-criteria in a given timehorizon.

Alternatives Sub-criteria


horizons. The total percentage of sustainability criteria driver atthe supplier level is estimated based on the summation of the per-centages of the sustainability criteria driver for each respectivesupplier within a specific time period.

S1 S2 S3 Sn

A1 B11 B12 B13 B1n

A2 B21 B22 B23 B2n

Am Bm1 Bm2 Bm3 Bmn

3.1.1. Sustainability factorsThe sustainability factors include economic, environmental and

social factors. A deduction process was employed to ascertain therelevant sustainability sub-criteria. Thus, economic factors(Amindoust et al., 2012; Chang et al., 2011; Govindan,Khodaverdi, & Jafarian, 2013; Ho, Xu, & Dey, 2010; Lee, Kang,Hsu, & Hung, 2009; Yeh & Chuang, 2010) include quality; environ-mental factors (Amin & Zhang, 2012; Bai & Sarkis, 2009; Kuo,Wang, & Tien, 2010; Shen, Olfat, Govindan, Khodaverdi, & Diabat,2013; Zhu & Sarkis, 2004) include environmental competencies(EC) and green design (GD) while social factors (Keskin, Ihan, &Ozkan, 2010; Kuo et al., 2010) include respect for policy (RFP),information disclosure (ID), and worker’s safety (WS).

Green design involves design process to enable ease of disas-sembling and recycling at the product end of life. In the case study,GD is embarked upon due to increased waste disposal costs andenvironmental legislation. Environmental competences incorpo-rate all knowledge and skills required for the effective environ-mental management. In the case study, workers undergotrainings to acquire skills to improve their environmental compe-tencies. Disclosure of information about the economic effects ofmanufacturing activities on the environment has become a signif-icant concern in business management (Bewley & Li, 2000). In thecase study information on environmental performance are dis-closed to assess the costs of pollution control. Worker’s safety isemphasized to reduce workplace injuries/accidents and healthhazards which tend o adversely affect operating efficiencies.Quality encompasses maximizing productivity and efficiency atminimized defects. Respect for policy (RFP) incorporates the com-pany’s policies with respect to human and workplace rights. In thecase study, RFP is emphasized thus placing a check on child labor,and other vices that could lead to legal cases thereby destroyingcompany image and causing financial loss.

3.1.2. Linguistic scaleThis is defined as a qualitative scale used to collect evaluator’s

judgment. The fuzzy linguistic scale employed has linguistic termsof very weak, weak, medium, good and very good with scores of 1,2, 3, 4 and 5 and triangular fuzzy numbers of (0,0,0.25),(0,0.25,0.50), (0.25,0.50,0.75), (0.50,0.75,1.00) and(0.75,1.00,1.00) respectively.

3.1.3. AlternativesIn this model, an alternative was defined as any supplier in the

gear manufacturing company.The fuzzy linguistic scale for the study is shown in Table 1.To obtain the information on the supplier behavior with respect

to deducted sustainability sub-criteria in a time horizon, fuzzydesign questionnaires were administered and the fuzzy linguistic

Table 1Fuzzy linguistic scale.

Linguistic term Score Triangular fuzzy numbers

Very weak 1 (0,0,0.25)Weak 2 (0,0.25,0.50)Medium 3 (0.25,0.50, 0.75)Good 4 (0.50,0.75,1.00)Very good 5 (0.75,1.00, 1.00)

scale in Table 1 is applied to obtain the decision matrix with ele-ments Ai, Si and Bij, shown in Table 2.

(Ai, Si and Bij are sustainability sub-criteria, supplier alternativesand deducted supplier behavior with regards to the sustainabilityfactor respectively).

The data in Table 2 will serve as input data to the systemdynamics model that will be developed.

A diffuzification process known as Converting Fuzzy data intoCrisps Scores (CFCS) process is applied to diffuzify the fuzzy setinto crisp values. It is deemed to be more effective by researchersfor arriving at crisp values when compared to the centroid method(Gharakhani, 2012). A triangular fuzzy number can be shown asq = (a,b,c) and the triangular membership function ~u0q is definedas Eq. (1).

~u0qðyÞ ¼

0ðy�aÞ

if y < a

if a 6 y 6 bðb�aÞðc�yÞ

ðc � bÞif b 6 y 6 c

0 if y > c

8>>>>>><>>>>>>:

ð1Þ

CFCS is based on determination of fuzzy maximum and minimum ofthe fuzzy number range. According to the membership function ~u0q,the total score is calculated with the weighted average. Given that Urepresents a fuzzy set, the fuzzy evaluation is given by

qdij ¼ ðad

ij; bdij; c

dijÞ for decision makers d = (1, 2, . . . , n) for the degree

of influence of sub-criterion i on sub-criterion j. The CFCS methodinvolves a five-step algorithm described as follows (Orji & Wei,2014):

Step one: Normalization:

xanij ¼ ðan

ij �mincnijÞ=D

maxmin ð2Þ

xbnij ¼ ðb

nij �mincn

ijÞ=Dmaxmin ð3Þ

xcnij ¼ ðcn

ij �mincnijÞ=D

maxmin ð4Þ

where Dmaxmin ¼ maxan

ij �mincnij ð5Þ

Step two: Compute right (as) and left (cs) normalized values:

xasnij ¼ xan

ij=ð1þ xanij � xbn

ijÞ ð6Þ

xcsnij ¼ xbn

ij=ð1þ xbnij � xcn

ijÞ ð7Þ

Step three: Compute total normalized crisp values:

xnij ¼ ½xcsn

ijð1� xcsnijÞ þ xasn

ijX xasnij�=½1� xcsn

ij þ xasnij� ð8Þ

Step four: Compute crisp values:

unij ¼ mincn

ij þ xnijXDmax

min ð9Þ

Step five: Integrate crisp values:

uij ¼ 1=pðu1ij þ u2

ij þ � � � þ upijÞ ð10Þ


3.2. Multi-criteria decision making model (MCDM)

A multi-criteria decision making model based on fuzzy TOPSIS(Technique for Order Performance by Similarity to Ideal Solution)is presented to compare results from the system dynamics model.TOPSIS is based on the concept of the measures of distance of thealternatives from the positive and negative ideal solutions. It ismost preferred as a straightforward multi-criteria decision makingmodel and has been extensively applied in literature (Awasthiet al., 2010; Govindan et al., 2013; Khamseh & Mahmoodi, 2014;Wang et al., 2009; Wittstruck & Teuteberg, 2011).

The advantages of TOPSIS over other multi-criteria decisionmaking models include the following (Govindan et al., 2013):

� An unlimited range of criteria and performance attributes canbe included.� It allows explicit trade-offs and interactions among attributes.

More precisely, changes in one attribute can be compensatedfor in a direct or opposite manner by other attributes.� Preferential ranking of alternatives with numerical value that

provides a better understanding of differences and similaritiesbetween alternatives, whereas other multi-criteria decisionmaking techniques (such as the ELECTRE) only determine therank of each alternative.� TOPSIS avoids pair wise comparisons employed by other

multi-criteria decision making techniques and thus can beeffectively employed when dealing with a large number of cri-teria and alternatives.� TOPSIS is a straightforward and relatively simple computation

process with a systematic procedure.� TOPSIS has the fewest rank reversals when an alternative is

added or removed among the multi-criteria decision makingtechniques.

Typically, TOPSIS technique consists of the following steps:Step one: Develop a normalized decision matrix.The normalized decision matrix is developed with element Pij

which represents the normalized evaluation index for the alterna-tive suppliers as shown in Table 3.

Pij is computed as:

Pij ¼gij

ðg2ijÞ

0:5 ð11Þ

wheregij is the performance of each alternative with respect to eachcriterion.Step two: Calculate the weighted normalized decision matrix.The weighted normalized decision matrix Vij is calculated as

follows:

Vij ¼ Pij � aij ð12Þ

whereaij is the normalized weight of indexes.aij is calculated from the divergence through ej of each criterionas shown below:

Table 3Normalized decision matrix.

Alternatives Sub-criteria

S1 S2 S3 Sn

A1 P11 P12 P13 P1n

A2 P21 P22 P23 P2n

Am Pm1 Pm2 Pm3 Pmn

aij ¼ei

Rejð13Þ

The divergence through of each criterion is computed from theentropy measures Dj of the respective criterion as follows:

ej ¼ 1� Dj ð14Þ

The entropy measures are calculated from the normalized eval-uation index Pij for the alternative suppliers as shown:

Pij ¼ �KX

i

½Pij � ln Pij� ð15Þ

whereK is a constant and the inverse of the natural logarithm of thetotal number of supplier alternatives.Step three: Calculate the positive and negative ideal solutions.The ideal (Vj

+) and negative ideal (Vj�) solutions are determined

as:

Vþj ¼ fvþi . . . vþn g ¼ ½ðmax v ijji 2 I0Þ; ðmin v ijji 2 I00Þ� ð16Þ

V�j ¼ fv�i . . . v�n g ¼ ½ðmin v ijji 2 I0Þ; ðmax v ijji 2 I00Þ� ð17Þ

where I0 is associated with advantage criteria, and I00 is associatedwith cost criteria.

Step four: Compute the separation measures.The separation measures (dþi and d�i ) are computed using the

n-dimensional Euclidean distance for the alternatives as:

dþi ¼Xn

j¼1

ðv ij� vjÞ2( )0:5

ð18Þ

where i = 1, 2, . . . , m; vj (in Eq. (12)) = vj+

d�i ¼Xn

j¼1

ðv ij� vjÞ2( )0:5

ð19Þ

where i = 1, 2, . . . , m; vj (in Eq. (13)) = vj�.

Step five: Determine the relative closeness to the ideal solutionfor the supplier alternatives.

The relative closeness of the alternatives (aj) to the ideal (A⁄)solution is computed as:

Li ¼d�i

d�i þ dþi0 6 Li 6 1 ð20Þ

where I0 is associated with advantage criteria, and I00 is associatedwith cost criteria.

Finally, the alternative suppliers are ranked with respect totheir relative closeness to the ideal solution in order of preference.

4. Results and discussion

Data analysis was carried out using Microsoft EXCEL, MATLABand Vensim. Microsoft EXCEL was applied in the CFCS diffuzifica-tion process to convert fuzzy data sets in the time horizons to crispscores. MATLAB was applied in the fuzzy-TOPSIS approach todevelop the normalized and weighted normalized decision matri-ces. Vensim was applied to define the causal loop diagram, developthe systems dynamics model and run simulations in four scenarios.The four scenarios represent the different time horizons in thisstudy. Also Table 4 shows the triangular fuzzy numbers for sup-plier performance for one of the experts in the past period.

The triangular fuzzy numbers of experts were normalized usingCFCS process into crisp values as shown in Eqs. (2)–(5). Table 5shows the normalized triangular fuzzy numbers of an expert in aperiod.

Table 4Triangular fuzzy numbers of an expert in a period.

A1 A2 A3 A4

WS (0.25,0.50,0.75) (0.50,0.75,1.00) (0.25,0.50, 0.75) (0.50,0.75,1.00)GD (0.50,0.75,1.00) (0,0,0.25) (0,0,0.25) (0.75,1.00, 1.00)Quality (0,0.25,0.50) (0.50,0.75,1.00) (0.75,1.00, 1.00) (0,0.25,0.50)RFP (0.50,0.75,1.00) (0.75,1.00,1.00) (0.50,0.75,1.00) (0.50,0.75,1.00)ID (0.25,0.50,0.75) (0.50,0.75,1.00) (0.25,0.50, 0.75) (0.25,0.50, 0.75)EC (0.50,0.75,1.00) (0.25,0.50,0.75) (0.75,1.00, 1.00) (0,0,0.25)

Table 5Normalized triangular fuzzy numbers of an expert in a period.

A1 A2 A3 A4

WS (0,0.33,0.66) (0.33,0.66,1) (0,0.33,0.66) (0.33,0.66,1)GD (0.5,0.75,1) (0,0,0.25) (0,0,0.25) (0.75,1.00, 1.00)Quality (0,0.25,0.5) (0.50,0.75,1.00) (0.75,1.00,1.00) (0,0.25,0.50)RFP (0,0.5,1.5) (1.5,1.00,1.00) (0,0.5,1) (0,0.5,1)ID (0,0.33,0.66) (0.33,0.66,1) (0,0.33,0.66) (0,0.33,0.66)EC (0.50,0.75,1.00) (0.25,0.50,0.75) (0.75,1.00,1.00) (0,0,0.25)

Table 6Computed right and left normalized values of triangular fuzzy numbers of an expert in a period.

A1 A2 A3 A4

WS (0.24,0.33,0.49) (0.66,0.66,0.49) (0.24,0.33,0.49) (0.66,0.66,0.49)GD (0.6,0.75,0.8) (0,0,0.2) (0,0,0.2) (0.8,1,1)Quality (0.2,0.25,0.33) (0.6,0.75,0.8) (0.8,1,0.8) (0.2,0.25,0.33)RFP (0.6,0.5,0.75) (0.2,1.00,0.66) (0.33,0.5,0.5) (0.33,0.5,0.5)ID (0.24,0.33,0.66) (0.66,0.66,0.49) (0.24,0.33,0.66) (0.24,0.33,0.66)EC (0.6,0.75,0.8) (0.4,0.50,0.5) (0.8,1,0.8) (0,0,0.2)

Table 7Total normalized crisp values of fuzzy numbers of an expert in a period.

A1 A2 A3 A4

WS 0.338 0.559 0.338 0.559GD 0.733 0.333 0.333 0.966Quality 0.237 0.733 0.800 0.237RFP 0.979 0.407 0.210 0.210ID 0.435 0.559 0.435 0.435EC 0.733 0.454 0.800 0.333

Table 8Supplier performance with respect to criteria in past period.

A1 A2 A3 A4


Table 9Supplier performances with respect to criteria in future horizon.

A1 A2 A3 A4



The left and right side normalized values were calculated usingEqs. (6) and (7) as shown in Table 6.

Table 7 shows the normalized crisp values of an expert within aperiod which was calculated using Eq. (8). The total normalizedcrisp values were calculated using Eq. (9). Then the crisp valueswere integrated using Eq. (10). The same process was employedto compute the total normalized crisp values and integrate thecrisp values of supplier behavior for all the experts.

A simple average method was then employed to estimate theaverage value for supplier behavior in each time horizon. Table 8shows the supplier performances with regards specific sustainabil-ity criteria in the past period.

The same process can be repeated in the current and futuretime horizons to obtain supplier behavior in the current and futureperiod respectively. Table 9 states the predicted data on perfor-mances of suppliers with regards to sustainability criteria in theperiod from June 2014 to June 2016.

Fig. 3 shows a Systems Dynamics Sustainable Supplier SelectionModel in Vensim.

The systems dynamics model presented shows the variables forfour suppliers with respect to six sustainability criteria namely

work safety, environmental competencies, respect for policy, infor-mation disclosure, green design and quality. The similar structurecould be replicated for additional suppliers and sustainabilitycriteria.

For a proper understanding of how supplier behavior evolveswith time, a comparative analysis of supplier performance withrespect to deducted sustainability criteria, in the past, currentand future time horizon has been carried out by simulations inVensim. Simulation is a method which contributes to innovationprocess by facilitation of virtual experimentation. Data presented

Fig. 3. Systems dynamics model for sustainable supplier selection.


in Tables 4–9 were used as inputs for the simulation runs. The sup-plier performance in the future horizon was captured using predic-tions by experts based on the investment trends of the suppliersfor the future period. Many decisions in companies are strategicdecisions for the future, and these have been criticized for not con-sidering future predictions, thus resulting in unrealistic decisions(De Boer, Labro, & Morlacchi, 2001; Ho et al., 2010). Fig. 4 showsthe behavior of a sustainability criterion for a particular supplier

in the different time horizons. As shown, the performance percent-age of a sustainability criterion remains constant in each time hori-zon. This behavior of a criterion is due to the underlyingassumption in this model that total budget of a supplier is fixedand can only be altered at the beginning of each time horizon.Hence, the percentage of each sustainability criterion is constantthroughout the specific time horizon. Similar behavior was repli-cated by other sustainability criteria irrespective of supplier.

.4

.35

.3

.25

.2

4

3

2

1

0 2 4 6 8 10 12 14 16 18 20 22 24Time (Month)

Perc

ent

"%of GD performance driver at the supplier level for supplier number1" : Average 1 1 1 1 1"%of GD performance driver at the supplier level for supplier number1" : Future period 2 2 2 2"%of GD performance driver at the supplier level for supplier number1" : Current period 3 3 3 3"%of GD performance driver at the supplier level for supplier number1" : Past period 4 4 4 4

Fig. 4. Vensim simulation results of a sustainability criterion in the different timehorizons.


Fig. 5 shows the behavior of rate of investment in sustainabilityfactors by a supplier.

The rate of investment in sustainability factors increases contin-uously for the suppliers alternatives because of continuous empha-sis on sustainable development as shown in Fig. 5. Only thebehaviors of supplier alternatives involved in sustainable develop-ment in their firms were analyzed in this work. The great increasein the rate of investment in sustainability factors in the currentperiod is caused by an increase in available total budget. The pre-dicted data on supplier performances with respect to sustainabilitycriteria in the future period is similar to that of the past period,hence the similarity in performance curve in the both periods.The percentage performance in sustainability criteria can be calcu-lated by dividing the rate of investment in sustainability factors of

70,000

52,500

35,000

17,500

04 4 4 4 4 4 4 4 4 43 3 3 3 3 3 3

3

3

3

3

2 2 2 2 2 2 2 2 2 2 21 1 1 1 1 1 1 1 1 1

1

Time (Month)

CN

Y

rate of investment in sustainability factors by supplier number1 : Average 1 1 1 1 1rate of investment in sustainability factors by supplier number1 : Future period 2 2 2 2 2rate of investment in sustainability factors by supplier number1 : Current period 3 3 3 3 3rate of investment in sustainability factors by supplier number1 : Past period 4 4 4 4 4

rararara

20,000

15,000

10,000

5000

04 4 4 4 4 4 4 4 4 43 3 3 3 3 3 3

3

3

3

3

2 2 2 2 2 2 2 2 2 2 2

1 1 1 1 1 1 1 1 1 11

Time (Month)

CN

Y


0 2 4 6 8 10 12 14 16 18 20 22 24

0 2 4 6 8 10 12 14 16 18 20 22 24

(a)

(c)

Fig. 5. Vensim simulation results of rate of investment in sustainability factors for (a) sup

the different suppliers by their respective total sustainability per-formance. Given the information on four suppliers (supplier num-ber 1, supplier number 2, supplier number 3 and supplier number4), their percentage performance of sustainability criteria are 36%,40%, 21% and 25% respectively.

Fig. 6 shows the behavior of a supplier with respect to sustain-ability criteria in the past period, current period, future period andthe average sustainable supplier behavior for the three periods.The average supplier behavior for the three periods was computedfor the different suppliers using the simple average method.

As shown in Fig. 6, total performance of a sustainable suppliergrows exponentially, because of the increase in the rate of invest-ment in sustainability factors by the supplier. This leads to highertotal performance in the different time horizons. Also, the growthrate of the total performance of suppliers outruns their rate ofinvestment in sustainability after about 12 months. With respectto the average of the total performance of the different suppliersin the past, current and future periods, the best possible sustain-able supplier can be selected.

Table 10 shows the ranking of suppliers in the different timeperiods.

The ranking in the average column as shown in Table 10 wasused as the final ranking of the suppliers. Thus the supplier number2 ranks the highest with regards to the relevant sustainability cri-teria. This is followed by supplier number 1, then supplier number4. The least alternative with respect to the relevant sustainabilitycriteria is supplier number 3.

A comparison analysis between the systems dynamics modeland the multi-criteria decision making model (MCDM) is proposedto provide insights on the viability of the systems dynamic model.The data inputs for the systems dynamics model is also used forthe MCDM. Table 11 shows the normalized decision matrix inthe current horizon.

200,000

150,000

100,000

50,000

04 4 4 4 4 4 4 4 4 43 3 3 3 3 3 3

3

3

3

2 2 2 2 2 2 2 2 2 2 21 1 1 1 1 1 1 1 1 1

1

Time (Month)

CN

Y

te of investment in sustainability factors by supplier number2 : Average 1 1 1 1 1te of investment in sustainability factors by supplier number2 : Future period 2 2 2 2 2te of investment in sustainability factors by supplier number2 : Current period 3 3 3 3 3te of investment in sustainability factors by supplier number2 : Past period 4 4 4 4 4

30,000

22,500

15,000

7500

04 4 4 4 4 4 4 4 4 43 3 3 3 3 3 3

3

3

3

3

2 2 2 2 2 2 2 2 2 22

1 1 1 1 1 1 1 1 1 11

0 2 4 6 8 10 12 14 16 18 20 22 24

0 2 4 6 8 10 12 14 16 18 20 22 24

Time (Month)

CN

Y


(b)

(d)

plier number 1 (b) supplier number 2 (c) supplier number 3 (d) supplier number 4.

200,000

150,000

100,000

50,000

04 4 4 4 4 4 4 4 4 4

3 3 3 3 3 33

3

3

3

2 2 2 2 2 2 2 2 2 22

1 1 1 1 1 1 1 1 11

1

Time (Month)

CN

Y

total performance of supplier number1 : Average 1 1 1 1 1 1total performance of supplier number1 : Future period 2 2 2 2 2 2total performance of supplier number1 : Current period 3 3 3 3 3total performance of supplier number1 : Past period 4 4 4 4 4

400,000

300,000

200,000

100,000

04 4 4 4 4 4 4 4 4 43 3 3 3 3 3

33

3

3

2 2 2 2 2 2 2 2 2 2 21 1 1 1 1 1 1 1 1 1

1

Time (Month)

CN

Y


60,000

45,000

30,000

15,000

04 4 4 4 4 4 4 4

44

3 3 3 3 3 33

3

3

3

3

2 2 2 2 2 2 2 2 22

2

1 1 1 1 1 1 1 1 11

1

Time (Month)

CN

Y


80,000

60,000

40,000

20,000

04 4 4 4 4 4 4 4

44

3 3 3 3 3 3 3

3

3

3

3

2 2 2 2 2 2 2 2 22

2

1 1 1 1 1 1 1 1 11

1

0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24

0 2 4 6 8 10 12 14 16 18 20 22 24 0 2 4 6 8 10 12 14 16 18 20 22 24Time (Month)

CN

Y


(a) (b)

(c) (d)

Fig. 6. Vensim simulation results of total sustainability performance for (a) supplier number 1 (b) supplier number 2 (c) supplier number 3 (d) supplier number 4.

Table 10Ranking of suppliers in different time periods.

Past Current Future Average

A1 3 4 1 2A2 2 2 2 1A3 1 3 4 4A4 4 1 3 3

Table 11Normalized decision matrix in current time period.

A1 A2 A3 A4


Table 12Average scores for the weighted normalized decision matrix.

GD WS Quality RFP ID EC

A1 0.5723 0.8977 0.6547 1.01 0.6949 0.5256A2 0.8412 0.4975 1.0536 1.01 0.6949 0.7147A3 1.0386 0.2312 0.8105 0.5156 0.5129 0.5256A4 0.7751 0.6826 0.7088 0.5434 0.4799 0.712

Table 13Total performance and ranking of suppliers in current period.

d�i dþi dþi þ d�i Li Ranking

A1 0.7218 0.732 1.4538 0.379459 4A2 0.631 0.4066 1.0376 0.608134 2A3 0.805 0.705 1.51 0.5331 3A4 0.4744 0.7758 1.2502 0.9692 1


The weighted normalized decision matrix as already stated inEq. (12) was calculated for the current period. Also the weightednormalized decision matrix was computed as shown in Table 12.

The positive and negative ideal solutions including were deter-mined using Eqs. (16) and (17) for the current time horizons. Alsothe separation measures (dþi and d�i ) were computed using then-dimensional Euclidean distance for the alternatives in Eqs. (18)and (19) for the suppliers in the current time horizon.

Finally, the relative closeness to the ideal solutions Li (TOPSISindex) which is the total performance of the respective supplierswere calculated using Eq. (20) in the respective time periods. Thecalculated separation measures and relative closeness to the idealsolutions in the current time horizon are shown in Table 13.

From the ranking shown in Table 13, supplier number 4 (A4) isidentified as the best sustainable supplier. This is similar to theresults of the systems dynamics model as shown in Table 10 forthe current period. The calculated relative closeness to the idealsolutions Li (TOPSIS index) for the supplier alternatives in theaverage time are shown in Table 14.

The results of the multi-criteria decision making model in theaverage period shown in Table 14 differ from those of the systemsdynamics model in the average time horizon shown in Table 10.Multi-criteria decision making model are able to give reliableresults in the current time horizon. They do not provide adequateinformation on how supplier behavior evolves with time. There isno doubt that the validity of the result is rather limited, specificallywhen the values change over time and the decision matrix is notfixed or static as in sustainable supplier selection problems.

Table 14Total performance and ranking of suppliers in average time.

d�i dþi dþi þ d�i Li Ranking

A1 0.805 0.705 1.51 0.533113 3A2 0.4744 0.7758 1.2502 0.379459 4A3 0.7218 0.732 1.4538 0.9692 1A4 0.631 0.4066 1.0376 0.608134 2


Hence, in this work, the systems dynamics model is advocated toprovide reliable information on how supplier behavior evolveswith time.

5. Conclusion

Sustainable supplier selection is a very critical agent to ensuringthe profitability and survival of a company. However, a major prob-lem while implementing sustainable supplier selection is how toinsure that suppliers maintain their status for a long period.Hence, this study presented an integrated fuzzy logic and systemsdynamics modeling approach for sustainable supplier performancemonitoring and selection in supply chain systems capable of insur-ing suppliers maintain their status for a long period. Supplier per-formance with respect to deducted sustainability criteria has beenestimated by experts using fuzzy questionnaires to serve as inputdata to systems dynamics model. Also a systems dynamics diagramin Vensim software has been developed and simulation runscarried out in four different time horizons namely, past, current,future and average.

The model provides an understanding of how supplier behaviorevolves with time and selected the best possible sustainablesupplier. A multi-criteria decision making model based on fuzzyTOPSIS (Technique for Order Performance by Similarity to IdealSolution) was presented to compare results from the developedsystems dynamics model. Vensim simulation results show thatan increase in the rate of investment in sustainability factors alsocauses an exponential increase in total performance of suppliers.The results of the MCDM only show accurate supplier performancein the current time horizon. This is because the basic assumption inapplying decision making models is that both criteria and alterna-tives are fixed a priori and that decision occurs only once i.e., nospatial or temporal considerations are included. Specifically whenthe values change over time and the decision matrix is not fixedor static as in sustainable supplier selection problems, the validityof the results are rather limited. Thus, MCDM does not provideadequate information on how supplier behavior evolves with time.Rather, systems dynamics model is advocated.

The systems dynamics modeling approach presented in thisstudy can be applied to any green manufacturing environmentregardless of the number of alternatives and relevant sustainabilitycriteria. At the development of the systems dynamic model, timeand cost resources could be demanding, but once installed, use ofthe model becomes less demanding.

References

Aghajani, H., & Ahmadpour, M. (2011). Application of fuzzy TOPSIS for rankingsuppliers of supply chain in automobile manufacturing companies in Iran. FuzzyInformation and Engineering, 3(4), 433–444.

Akhtar, M. H., Simonovic, S. P., Wibe, P., MacGee, J., & Davies, J. (2011). An integratedsystem dynamics model for analyzing behavior of the social-energy-economic-climatic system: User’s manual. Report no: 076. Water resources researchreport.

Amin, S. H., & Zhang, G. (2012). An integrated model for closed-loop supply chainconfiguration and supplier selection: Multi-objective approach. Expert Systemswith Applications, 39, 6782–6791.

Amindoust, A., Ahmed, S., Saghafinia, A., & Bahreininejad, A. (2012). Sustainablesupplier selection: A ranking model based on fuzzy inference system. AppliedSoft Computing, 12(6), 1668–1677.

Awasthi, A., Chauhan, S. S., & Goyal, S. K. (2010). A fuzzy multi criteria approach forevaluating environmental performance of suppliers. International Journal ofProduction Economics, 126(2), 370–378.

Azadnia, A. H., Saman, M. Z. M., & Wong, K. Y. (2015). Sustainable supplier selectionand order lot-sizing: An integrated multi-objective decision-making process.International Journal of Production Research, 53(2), 383–408.

Bai, C., & Sarkis, J. (2009). Supplier selection and sustainability: A grey rough setevaluation. Working paper no. 2009-05.

Bayrak, M. Y., Celebi, N., & Taskin, H. (2007). A fuzzy approach method for supplierselection. Production Planning and Control: The Management of Operations, 18(1),54–63.

Belhajali, I., & Hachicha, W. (2013). System dynamics simulation to determine safetystock for a single-stage inventory system. IEEE. http://dx.doi.org/10.1109/ICAdLT.2013.6568507.

Bevilacqua, M., & Petroni, A. (2002). From traditional purchasing to suppliermanagement: A fuzzy logic-based approach to supplier selection. InternationalJournal of Logistics Research and Applications: A Leading Journal of Supply ChainManagement, 5(3), 235–255.

Bewley, K., & Li, Y. (2000). Disclosure of environmental information by Canadianmanufacturing companies: A voluntary disclosure perspective. Advances inEnvironmental Accounting & Management, 1, 201–226.

Bottani, E., & Rizzi, A. (2008). An adapted multi-criteria approach to suppliers andproducts selection—An application oriented to lead-time reduction.International Journal of Production Economics, 111, 763–781.

Briano, E., Caballini, C., Giribone, P., & Revetria, R. (2010). Using system dynamics forshort life cycle supply chain evaluation. In Proceedings of winter simulationconference.

Büyüközkan, G., & Çifçi, G. (2011). A novel fuzzy multi-criteria decision frameworkfor sustainable supplier selection with incomplete information. Computers inIndustry, 62(2), 164–174.

Campanella, G., Pereira, A., Ribeiro, R. A., & Varela, L. R. (2012). Collaborative dynamicdecision making: A case study from B2B supplier selection. In J. E. Hernández, P.Zarate, F. Dargam, B. Delibasic, S. Liu, & R. Ribeiro (Eds.), Decision support systems –Collaborative models and approaches in real environments (pp. 88–102). Berlin:Springer. http://dx.doi.org/10.1007/978-3-642-32191-7_7.

Campanella, G., & Ribeiro, R. A. (2011). A framework for dynamic multiple criteriadecision making. Decision Support Systems, 52(1), 52–60. http://dx.doi.org/10.1016/j.dss.2011.05.003.

Chang, B., Chang, C.-W., & Wu, C.-H. (2011). Fuzzy DEMATEL method for developingsupplier selection criteria. Expert Systems with Applications: An InternationalJournal, 38(3), 1850–1858.

Chen, C. M. (2009). A fuzzy-based decision-support model for rebuy procurement.International Journal of Production Economics, 122, 714–724.

De Boer, L., Labro, E., & Morlacchi, P. (2001). A review of methods supportingsupplier selection. European Journal of Purchasing & Supply Management, 7,75–89.

Elsawah, S., Haase, D., Delden, H., Pierce, S., Elmahdi, A., Voinovo, A. A., et al. (2012).Using system dynamics for environmental modelling: Lessons learnt from sixcase studies. In International environmental modelling and software society(iEMSs). International congress on environmental modelling and softwaremanaging resources of a limited planet.

Foerstl, K., Reuter, C., Hartmann, E., & Blome, C. (2010). Managing suppliersustainability risks in a dynamically changing environment—Sustainablesupplier management in the chemical industry. Journal of Purchasing andSupply Management, 16(2), 118–130.

Ford, A. (1999). Modeling the environment (2nd ed.). Island Press.Forrester, J. (1971). World dynamics. Cambridge, MA: Wright-Allen Press.Gharakhani, D. (2012). The evaluation of supplier selection criteria by fuzzy

DEMATEL method. Journal of Basic and Applied Scientific Research, 2(4),3215–3224.

Govindan, K., Khodaverdi, R., & Jafarian, A. (2013). A fuzzy multi criteria approachfor measuring sustainability performance of a supplier based on triple bottomline approach. Journal of Cleaner Production, 47, 345–354.

Haller, R., Emberger, G., Mayerthaler, A. (2008). A system dynamics approach tomodel land-use/transport interactions on the national level. In REAL CORP 008proceedings. <www.corp.at>.

Heyer, R. (2004). Understanding soft operations research: The methods, theirapplication and its future in the defence setting. Edinburgh South Australia:DSTO Information Sciences Laboratory (working paper).

Ho, W., Xu, X., & Dey, P. K. (2010). Multi-criteria decision making approaches forsupplier evaluation and selection: A literature review. European Journal ofOperational Research, 202, 16–24.

Jauhar, S. K., Pant, M., & Abraham, A. (2014). A novel approach for sustainablesupplier selection using differential evolution: A case on pulp and paperindustry. Intelligent data analysis and its applications (Vol. II, pp. 105–117).Springer International Publishing.

Javad, J., Rita, A., & Leonilde, R. (2014). Dynamic MCDM with future knowledge forsupplier selection. Journal of Decision Systems, Knowledge-Based Decision Systems,23(3).

Kahraman, C., Cebeci, U., & Ulukan, Z. (2003). Multi-criteria supplier selection usingfuzzy AHP. Logistics Information Management, 16(6), 382–394.

Keskin, G. A., Ihan, S., & Ozkan, C. (2010). The fuzzy art algorithm: A categorizationmethod for supplier evaluation and selection. Expert Systems with Application,37, 1235–1240. http://dx.doi.org/10.1016/j.eswa.2009.06.004.

http://refhub.elsevier.com/S0360-8352(15)00278-8/h0005


















http://dx.doi.org/10.1109/ICAdLT.2013.6568507

http://dx.doi.org/10.1109/ICAdLT.2013.6568507














http://dx.doi.org/10.1007/978-3-642-32191-7_7

http://dx.doi.org/10.1016/j.dss.2011.05.003

http://dx.doi.org/10.1016/j.dss.2011.05.003





















http://www.corp.at













http://dx.doi.org/10.1016/j.eswa.2009.06.004


Khamseh, A. A., & Mahmoodi, M. (2014). A new fuzzy TOPSIS-TODIM hybrid methodfor green supplier selection using fuzzy time function. Advances in FuzzySystems. http://dx.doi.org/10.1155/2014/841405.

Khatie, A. H., Bulgak, A., & Segovia, J. (2010). Advanced decision support tool byintegrating activity-based costing and management to systems dynamics. IEEE.

Kuo, R., Wang, Y., & Tien, F. (2010). Integration of artificial neutral network andMADA methods for green supplier selection. Journal of Cleaner Production,18(12), 1161–1170. http://dx.doi.org/10.1016/j.jclepro.2010.03.020.

Lan, T.-S., Lan, Y.-H., Chen, K.-L., Chen, P.-C., & Lin, W.-C. (2013). A study ofdeveloping a system dynamics model for the learning effectiveness evaluation.Mathematical Problems in Engineering. http://dx.doi.org/10.1155/2013/298621.

Lee, A. H. I., Chen, W.-C., & Chang, C.-J. (2008). A fuzzy AHP and BSC approach forevaluating performance of IT department in the manufacturing industry inTaiwan. Expert Systems with Applications (34), 96–107.

Lee, A. H. I., Kang, H.-Y., Hsu, C.-F., & Hung, H.-C. (2009). A green supplier selectionmodel for high-tech industry. Expert Systems with Applications, 36, 7917–7927.http://dx.doi.org/10.1016/j.eswa.2008.11.052.

Mani, V., Agrawal, R., & Sharma, V. (2014). Supplier selection using socialsustainability: AHP based approach in India. International StrategicManagement Review, 2(2), 98–112.

Meadows, J., Randers, D. L., & Meadows (1993). Beyond the limits: Confronting globalcollapse, envisioning a sustainable future. Chelsea Green Publishing Company.

Meadows, J., Randers, D. L., & Meadows (2004). Limits to growth: The 30-year update.Chelsea Green.

Onat, N., Egilmez, G., & Tatari, O. (2014). Towards greening the U.S. residentialbuilding stock: A systems dynamics approach. Building and Environment, 78,68–80.

Ordoobadi, S. (2009). Development of a supplier selection model using fuzzy logic.Supply Chain Management: An International Journal, 14(4), 314–327.

Orji, I. M. J., & Wei, S. (2014). A decision support tool for sustainable supplierselection in manufacturing firms. Journal of Industrial Engineering andManagement, 7(5), 1293–1315.

Qiang, X., Hui, L., & Xiao-dong, Q. (2013). System dynamics simulation model for theelectronic commerce credit risk mechanism research. International Journal ofComputer Science Issues, 10(2).

Randers, J. (2000). From limits to growth to sustainable development or SD(sustainable development) in a SD (system dynamics) perspective. SystemDynamics Review, 16, 213–224.

Shen, L., Olfat, L., Govindan, K., Khodaverdi, R., & Diabat, A. (2013). A fuzzy multi-criteria approach for evaluating green supplier’s performance in green supplychain with linguistic preferences. Resources, Conservation and Recycling, 74,170–179.

Tako, A., & Robinson, S. (2012). The application of discrete event simulation andsystem dynamics in the logistics and supply chain context. Decision SupportSystems, 52, 802–815.

Trappey, A., Trappey, C., Hsiao, C., Ou, J., & Chang, C. T. (2012). System dynamicsmodeling of product carbon footprint life cycles for collaborative green supplychains. International Journal of Computer Integrated Manufacturing, 25, 934–945.

Vashiranwongpinyo, P., (2010). A systems dynamics model to analyze behavior ofmanufacturing in supply chain. In Proceedings of the 2nd RMUTP internationalconference.

Verdecho, M. J., Alfaro-Siaz, J. J., & Rodríguez-Rodríguez, R. (2010). An approach toselect suppliers for sustainable collaborative networks. In IFIP internationalfederation for information processing (pp. 304–311).

Wang, J.-W., Cheng, C.-H., & Huang, K.-C. (2009). Fuzzy hierarchical TOPSIS forsupplier selection. Applied Soft Computing, 9(1), 377–386.

Wareef, A. N. (2013). A systems dynamics simulation model for forecasting energydemand in Pueblo County. MSc. Dissertation. Industrial and Systems EngineeringProgram, Colorado State University-Pueblo.

Wittstruck, D., & Teuteberg, F. (2011). Governance and sustainability in informationsystems. Managing the transfer and diffusion of IT. In IFIP advances ininformation and communication technology (Vol. 366, pp 45–69).

Wu, M., & Liu, Z. (2011). The supplier selection application based on two methods:VIKOR algorithm with entropy method and Fuzzy TOPSIS with vague setsmethod. International Journal of Management Science and EngineeringManagement, 6(2), 109–115.

Yeh, W. C., & Chuang, M. C. (2010). Using multi-objective genetic algorithm forpartner selection in green supply chain problems. Expert System Application, 38,4244–4253. http://dx.doi.org/10.1016/j.eswa.2010.09.091.

Zhang, D. W., Hamid, A., Bakar, A., & Thoo, A. C. (2014). Sustainable supplierselection: An international comparative literature review for futureinvestigation. Applied Mechanics and Materials, 525, 787–790.

Zhu, Q., & Sarkis, J. (2004). Relationships between operational practices andperformance among early adopters of green supply chain managementpractices in Chinese manufacturing enterprises. Journal of OperationsManagement, 22, 265–289.

http://dx.doi.org/10.1155/2014/841405

http://dx.doi.org/10.1016/j.jclepro.2010.03.020

http://dx.doi.org/10.1155/2013/298621
















































