Costumer Satisfaction 4

Operations Research Perspectives 6 (2019) 100113
Contents lists available at ScienceDirect
Operations Research Perspectives

journal homepage: www.elsevier.com/locate/orp
An integrated optimization model for selection of sustainable suppliers T

based on customers’ expectations
Venkata SS Yadavallia,⁎, Jyoti Dhingra Darbarib, Nidhi Bhayanac, P.C. Jhac, Vernika Agarwalc
a
Department of Industrial & Systems Engineering, University of Pretoria, Pretoria, South Africa
b
Department of Mathematics, Lady Shri Ram College, University of Delhi, Delhi, India
c
Department of Operational Research, University of Delhi, Delhi, India
ARTICLE INFO ABSTRACT
Keywords: Due to intensification of market competence and customers increasing socio-environmental concerns, sustain-
Customer sustainable expectations ability has become a focal point for most organizations. Thus incorporation of customer sustainable expectations
FTOPSIS into the conventional supplier selection process has become necessary for manufacturers to sustain in the
Z-numbers competitive market. To cope with this issue, it is necessary to model the customers’ opinions and expectations
Supplier selection
into the strategic supplier selection and order allocation decision making, an area not delved upon much in the
Goal programming
literature. Therefore, the aim of the study is to develop an analytical model for manufacturing firm for selection
of suppliers based on customers’ expectations which are reflected as retailers’ expectation level and also for
bringing in financial and socio-environmental stability to the whole SC. The research methodology adopted
involves 1) use of fuzzy-TOPSIS technique using Z numbers for evaluation of suppliers based on traditional as
well as social environmental attributes which reflect customer's social and green expectations, 2) development of
a novel bi-objective mathematical model for selecting suppliers meeting the expectation levels, allocating orders
proportionate to the performance scores of the suppliers, rewarding or penalizing the suppliers as per their
sustainable performance, 3) Transformation of the bi-objective model into a weighted goal programming model
for attaining a trade-off solution between the objectives of minimization of cost and maximization of sustainable
value of purchase. To exemplify the efficacy of the model, case of electronics sector is presented. The findings of
the result suggest that penalty and reward value incorporated in the cost objective system is effective in en-
hancing the sustainable performance of the suppliers and can prove to be an effective strategy for DMs. The
study recommends that the proposed mathematical model can considerably aid the manufacturer in integrating
supplier selection and order allocation decisions while optimizing the performance of the customer oriented
supply chain.
1. Introduction must gear towards integrating sustainability aspects into their buying
behavior [20,43]. Clearly organizations can improve upon their SC
Due to globalization, trade liberalization, fierce competition and performance and expand their practical competence by incorporating
rising sustainability related needs of customers, managing supply chain sustainability dimensions into the supplier selection process with the
(SC) has become quite complex. The whole SC is generally considered focus on fulfilling customers’ expectations [44,59]. However, managing
as a pull system which is customer-centric and therefore firms cannot a sustainable SC becomes a strategic issue for an organization in order
neglect the environmental and social concerns of the customers as it can to meet goals in terms of economic competitiveness and value creation
result in economic loss in terms of decreased return volumes [39,66]. [41]. Therefore focal firms must focus on sustainable supplier selection
Since manufacturer is a supplier to the customers and a buyer to the as a strategic response to the customer needs as environmental and
supplier therefore consideration of customers’ transformed purchasing social assessment of suppliers can act as an enabler towards improve-
attitude brings additional complexities in the purchasing decisions. ment of customers’ satisfaction level [52]. This satisfaction level is in-
Further, customer satisfaction is highly dependent on the environ- fluenced by many different aspects that organizations need to take into
mental and social impact of the final products and services. Hence, account as vital parameters into the supplier selection process for the
organizations need to be responsive enough to customer needs and success of the SC. Although literature is abundant with sustainable
⁎
Corresponding author.
E-mail address: sarma.yadavalli@up.ac.za (V.S. Yadavalli).
https://doi.org/10.1016/j.orp.2019.100113
Received 22 October 2018; Received in revised form 3 May 2019; Accepted 7 May 2019
Available online 10 May 2019
2214-7160/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
V.S. Yadavalli, et al. Operations Research Perspectives 6 (2019) 100113
supplier selection models, however very few studies have investigated 2.1. Sustainable supplier evaluation and attributes
supplier selection and order allocation models based on customers’
requirements. Additionally, appropriate selection of evaluation attri- Evaluation and selection of suppliers is one of the essential pro-
bute for representing the conflicting sustainability aspects [49] and cesses for attaining an efficient SC performance. This is due to the
need for trade-off clearly calls for a multi- attribute decision analysis distinct role of suppliers at all stages of the SC [62]. Maintaining long-
[7]. Further, selecting best performing suppliers and determining the term partnerships with suppliers and using fewer reliable suppliers can
optimum purchasing allocations while minimising the SC cost, needs help the manufacturers in increasing the value of the SC [35]. Tradi-
development of an optimisation model. tionally, the manufacturers used to focus only on traditional factors
The identified gap necessitates the need to focus on how to model such as reduction in operational cost and delivery reliability [10]. Re-
the customers’ sustainable expectations into the supplier evaluation cent studies advocate that the financial competency of the supplier is
attributes and procure as per the environmental and social compliance also an essential attribute for the evaluation purpose [43]. Recently,
of suppliers. In this regard, the contribution of the study is development there has also been considerable discussion in the literature related to
of a novel customer-oriented supplier selection and order allocation SC the need for inclusion of environmental aspects in supplier selection
model that can aid decision makers (DMs) of a manufacturing firm in process. The environmental capabilities of the suppliers have direct
making crucial decisions regarding selection of suppliers based on influence on the environmental performance of the SC, thus, the lit-
customers’ expectations and also bringing in financial and socio-en- erature has focused on the importance of necessary ecological certifi-
vironmental stability to the whole SC. The key features of the proposed cations for the suppliers [18,19]. Kuo et al. [38] proposed an integrated
mathematical model are: multi- attribute method that evaluates the factors used to select a green
suppliers based on their performance. Hsu et al. [27] proposed a DE-
1) Evaluation of suppliers based on customers’ environmental and so- MATEL based approach to analyze the performance of the suppliers
cial expectations in a multi- attribute decision making environment, based on their carbon management issues. Kannan et al. [31] has dis-
which has not been addressed in the literature. The customer's ex- cussed the role of suppliers in helping firms in achieving maximum
pectation level is translated in terms of retailer's expectation level of ecological-economic benefits. Dobos and Vörösmarty [14] proposed a
the supplier's performance. data envelopment analysis (DEA) based supplier evaluation metho-
2) Development of a bi-objective mixed integer programming model dology for determining the influence of environmental attribute on
with the objectives of minimizing the total cost of the system and supplier selection process.
maximizing the SVP. The supplier performance evaluation weights Due to the growing need of sustainability issues, organizations have
are also incorporated in the objectives. also started focusing on integrating sustainability aspects in the supplier
3) Attain a trade-off between the conflicting objectives of the total cost evaluation process [43]. It is also acknowledged that incorporation of
of SC network and the Sustainable Value of Purchase (SVP), while sustainable aspects in supply chain performance measurements also
ensuring that better performing suppliers (if selected) are rewarded enhances the concern of business firms [4]. Bai and Sarkis [6] devel-
and poor performing suppliers (if selected) are penalised, an aspect oped a modelling approach to address the supplier selection problem in
not considered in sustainable supplier selection models. sustainable environment. Authors integrated grey theory and TOPSIS
methodology for the evaluation process. In another study, Govindan
The research methodology adopted for achieving the above objec- et al. [21] have evaluated the suppliers based on their sustainable
tives entails the following: Evaluation of the suppliers is carried using performance. Luthra et al. [43] has ranked ‘Environmental costs,’
fuzzy-TOPSIS with Z numbers incorporating multiple DMs’ opinions ‘Quality of product,’ ‘Price of product,’ ‘Occupational health and safety
while considering the degree of reliability on information on which the systems,’ and ‘Environmental competencies’ as top five sustainable
opinions are based. The performance evaluation weights obtained supplier selection attribute. However, introducing the three dimensions
through Z-TOPSIS are incorporated in the cost objective for calculating into supplier selection process embeds a new set of trade-offs in the
penalty and reward value for selected suppliers and in the SVP objective decision, complicating the decision making process with both qualita-
for allocating the purchase order as per the performance weights of the tive and quantitative factors. It is a challenge how to measure social
suppliers. Goal programming methodology is effectively adopted for attribute in supplier selection against the economic and environmental
obtaining a trade-off between the objectives of minimisation of cost and management attribute and how to identify trade-offs between these
maximisation of SVP. In order to validate the application of the pro- three. Generally it becomes difficult to determine social sustainability
posed integrated model, a case study of an Indian manufacturer is parameters for supplier evaluation as DMs themselves have little con-
considered in this study and managerial and practical implications of ceptual understanding of the social impact of their SC and how it can be
the study are presented. measured [17]. Recently, Bai et al. [3] proposed a decision framework
The rest of the paper is organized as follows: Brief review of lit- in order to evaluate and select socially responsible suppliers by in-
erature and research gap is discussed in the Section 2. Section 3 dis- tegrating grey-based approach with best-worst method and TODIM in
cusses the problem description and methodology is presented in the order to identify the weights of social attributes and suppliers respec-
Section 4. Section 5 provides the numerical illustration. Result discus- tively. They have validated the applicability of proposed framework by
sion, sensitivity analysis and managerial implications are presented in taking a case study of an Iranian manufacturing company.
Section 6 followed by the conclusion in Section 7. The growth of the focus of organization on their core competencies
has led to tremendous gain in supplier selection over the past few
2. Literature review decades [19]. In the research by Sarkis and Talluri [54], it was sug-
gested that the organizations have started focussing more on the buyer-
The paper is presenting a bi-objective mathematical model which supplier relationship due to changing market demands which require
integrates supplier selection and allocation based on customer ex- specific skills and capabilities. Due to this current shift in the consumer
pectations in terms of sustainable attributes into the model while op- environmental awareness, it has become necessary to adopt green po-
timizing cost of the firm and maximizing SVP. The literature review licies and practices in supplier selection [51]. Although there are many
presented in sub-Sections 2.1 to 2.3 gives the gist of the work done by studies concerned with developing supplier selection models but the
researchers in recent years on evaluation attributes for suppliers with literature lacks in studies that deal with fulfilling customers’ sustainable
respect to sustainable aspects, sustainable supplier selection models and expectations. The few studies include the following: Lin and Huang
order allocation incorporating customer expectations and multi-attri- [42] identified the factors influencing consumer choice behaviour with
bute analysis approach for sustainable supplier evaluation models. respect to green products. They have examined the significant
2
differences in choice behaviour and consumption values of customers Various authors have also considered penalty and reward system for
on environmental issue. The study by Tseng and Hung [59] proposed suppliers considering varied aspects. Starbird [58] discussed the impact
model related to service quality and identified the gap between con- of penalty and reward system along with inspection policy on cost
sumers’ expectations and their perception with respect to green pro- minimizing supplier. In an another study by Mathur and Shah [45]
ducts. Coskun et al. [13] developed a model based on the customers’ analysed the impact of penalty parameters on decisions related to
expectations of buying green products. Authors identified the de- suppliers’ capacity, SC efficiency and distribution of SC profit among
termination level of three types of customers which affect the green partners. Selviaridis and Norrman [56] explored various challenges
network. Gücdemir and Selim [23] integrated the two approaches- related to performance based contracts for logistics service provider.
customer relationship management and production planning and con- They have considered reward and risk sharing incentives based on the
trol. The aim of the integration was to optimally utilize the resources in attitude of customers and provider. Wang et al. [61] developed the
order to satisfy the customers’ needs. Yazdani et al. [63] utilized an penalty reward system to encourage recycling process in closed loop SC
integrated approach for the selection of green suppliers by taking en- (CLSC) in order to reduce the cost and improve waste electrical and
vironmental attributes. Authors first identified inter-relationships and electronic equipment collection. Authors analysed several CLSC settings
degree of relationships between attributes and customer requirements related to collection rate and sharing of responsibility of manufacturer
and finally prioritize the suppliers. and collector. Inderfurth and Clemens [28] showed the impact of risks
The present study aims to overcome this gap by considering social in the buyers’ and suppliers’ coordination having deterministic demand.
attributes along with traditional and environmental attributes in the The above studies reinforce that sustainable supplier selection and
evaluation process generally from the perspective of fulfilling custo- order allocation along with penalty and reward system can aid in im-
mers’ expectations level in terms of sustainable practices. The attributes proving the performance of suppliers. However, none of the above
are identified through an extensive literature survey and deliberations studies have integrated penalty-reward system in supplier selection and
with multiple DMs. order allocation models based on customer expectation level as done in
this study.
2.2. Sustainable supplier selection and order allocation
2.3. Multi-attribute analysis approach for sustainable supplier evaluation
In recent scenario, researchers are integrating two problems viz.
supplier selection and order allocation problem in sustainable en- The strategic decision of supplier selection requires the ability of the
vironment in order to reduce their costs and enhance the performance decision making body to consider multiple attributes to prioritize the
of their SC significantly [67]. Often it becomes difficult to meet either suppliers and arrive at a final decision [26]. This causes the supplier
total demand or the attribute constraints from a single supplier, then selection process to heavily reply on multiple attribute decision models
orders must be placed with multiple suppliers. Hammami et al. [24] (MADM) [19]. The study by Chai et al. [10], concludes that MADM
proposed a method for handling multiple suppliers under various cur- allows the DMs to simplify the supplier selection process while balan-
rency conditions depending on the supplier location. Scott et al. [55] cing a variety of conflicting attribute. There has been substantial dis-
addresses the supplier selection problem using hybrid AHP with Quality cussion in literature on the employment of variety of multi-attribute
Function Deployment and chance constrained optimization method for decision making (MADM) approaches in supplier selection [30]. These
selection of best suppliers and simultaneous order allocation. Often include analytical hierarchy problem (AHP) [15], analytic network
multiple supplier selection problems become more complex as the re- process (ANP) [25], technique for order of preference for similarity to
quirements of DMs’ increase and numerous objectives need to be sa- ideal solution (TOPSIS) [53], grey system theory [5], soft computing
tisfied simultaneously. This requires integration of multi-attribute techniques such as genetic algorithm [16], fuzzy logic [9], neural net-
techniques and multi-objective programming to handle conflicting at- work [33], etc. Various uncertainties such as imprecision, incomplete-
tributes and trade-off among objectives [30,48]. Mohammed et al. [47] ness of data, ambiguity in the DMs’ opinions also arise when dealing
integrated multi-attribute decision making and fuzzy multi-objective with social and environmental attribute [36]. In this regard, it has
optimization methodology for selecting sustainable supplier and allo- become essential to consider an effective method which can handle
cating orders by considering all the three pillars of sustainability. Va- these kind of uncertainties [34,40]. Recent studies addressing this issue
hidi et al. [60] proposed a two-stage mixed possibilistic-stochastic include the following: Amindoust et al. [1] handled the subjectivity of
programming approach for bi-objective optimization model for the se- DMs’ by using fuzzy inference system (FIS) for supplier selection pro-
lection of sustainable supplier and order allocation. Govindan and Si- blem. Shen et al. [57] proposed fuzzy multi attribute approach to
vakumar [20] suggested a two phase approach for selection of the best evaluated suppliers with respect to their environmental performance.
green supplier. The authors used fuzzy TOPSIS followed by multi-ob- Hashemi et al. [25] suggested an integrated analytic network process
jective linear programming to minimize cost, material rejection, late (ANP) and Grey relational analysis (GRA) model to deal with the sup-
delivery recycle waste and carbon emissions. Azadnia et al. [2] sug- plier selection uncertainties. Orji and Wei [50] integrated Fuzzy logic
gested an integrated approach of fuzzy AHP and Multi-objective Linear and system dynamics for sustainable supplier selection. Zhang and Xu
Programming (MOLP) for sustainable supplier selection and order al- [65] developed a technique based on consensus maximization and
location combined with multi-period multi-product lot-sizing problem. fuzzy TOPSIS which helps in solving the interval-valued intuitionistic
Moghaddam [46] proposed a multi-objective optimization problem fuzzy MADM problems. Hashemi et al. [25] integrated ANP and GRA
(MOOP) in order to select the performing supplier and identify the for selection of green suppliers.
number of parts and products in a CLSC network. Moghaddam et al. On analysing the literature review it can be seen that there is a lack
[46] developed a fuzzy multi objective model by considering the eco- of integrated methodologies which investigate the supplier selection
nomic risk and late deliveries for supplier selection. Scott et al. [55] problem from the account of customers’ sustainable expectations in a
proposed an integrated approach using AHP- QFD combined with fuzzy environment. This study utilises modification of TOPSIS method
chance-constrained optimization algorithm for the selection of supplier based on Z-numbers, called Z-TOPSIS, which is effective in quantifying
and allocating optimal orders among them. Choudhary and Shankar the DMs’ opinions while taking into account their reliability, a well
[12] proposed a multi-objective integer linear programming model suited technique for the considered problem of supplier evaluation
related for issues such as selection of supplier, lot sizing and selection of based on multi-stakeholders’ perspectives. Further, very few authors
carrier. Jadidi et al. [29] developed the MOOP for selecting the best have developed optimisation models integrating supplier selection and
supplier in which three objectives were considered includes price, order allocation decisions while optimising the sustainable performance
number of rejections and lead time. of the SC network as done in this study.
3
2.4. Research contribution manufacturing plant (MP) and the finished product is then send to the
zonal distributors. The zonal distributors have designated distribution
The research gap identified in Sections 2.1–2.3 reinforces that, in- centres (ZDCs) where the inventory is stored and distributed to the
tegrated supplier selection and order allocation SC models along with various retail centres (RCs) in their zone as per the customer demand.
penalty-reward system to mend upon the customer satisfaction level are The distribution of products between various facilities is managed by
not found in the literature. This forms the basis of the following re- the 3PLPs.
search contribution of the study: In this system, it is the responsibility of the retailer to establish retail
relationships with the customers. Thus, the demand of each RC varies
• The inclusion of social attributes along with traditional and en- based on the demand of the customers. The retailer is responsible for
vironmental aspects in the evaluation process for fulfilling the sus- forecasting the demand of the product and identifying the satisfaction
tainable expectations of customers has not been discussed earlier. level of the customers which is then communicated to the manu-
Moreover, the novel attributes chosen in the study so that the sup- facturer. The company aims to achieve their economic targets while
pliers can be assessed on aspects which reflect customer's expecta- satisfying the needs and wants of customers. Since the customers now
tions. want environmental and social benefits along with economic viability
• The study utilises Z-TOPSIS to evaluate suppliers based on custo- of the products they buy, therefore these requirements thereby, become
mers’ environmental and social expectations in a multi- attribute the basis for the sustainable targets for the company. Therefore the
fuzzy decision making environment, which has not been addressed company wants to enhance its SC performance by including sustainable
in the literature. In addition to this, the opinions of the decision aspects into their decision making process. However, meeting the sus-
makers are based on their experience and access to the information tainable expectation of the customers requires strategic rethinking of
which may be uncertain or incomplete. Hence quantifying their the SC. To begin with, the company is planning to design two new
opinions as Z numbers which takes into consideration the degree of variants of toasters, based on the economic viability and green ex-
reliability of information is very critical as is done in this study. pectations of customers. The next essential decision that the company
• A novel bi-objective mathematical model is developed which in- needs to make is to understand which suppliers to work with who are
corporates suppliers’ assessment scores for selecting better per- flexible enough towards the changing requirements and are able to
forming sustainable supplier and allocating orders based on the meet their expectation level. Flexibility of the suppliers is required in
customers’ expectation level along with penalty and reward system terms of adopting new technologies, and inclination towards energy
while obtaining a trade-off solution between the objectives of conservation and social upliftment. Thus, collaboration is needed be-
minimizing SC cost and maximizing the SVP. tween the suppliers and manufacturer in order to improvise upon the
sustainable performance of the SC and increase the market share. This
This model enhances the customer satisfaction by establishing a clearly requires overhauling the supplier selection process and taking
strategic partnership with the selected supplier while simultaneously appropriate decisions at strategic and tactical level. At strategic level,
optimizing the cost and value of purchase. the need of company to evaluate and select the best performing supplier
on the basis of vital sustainable attribute while at tactical level, the
3. Problem description optimum order allocation must be identified. Since retailers are in di-
rect communication with the customers therefore they play a key role
The current study discusses the problem of a northern India based in identifying their needs. Hence retailers’ involvement in the supplier
home appliance manufacturing company. The company was started selection decision making process is crucial for choosing the attribute of
around three decades ago with only a single product and subsequently evaluation which can translate the needs of the customers and assessing
climbed to become one of India's foremost OEMs in Home Appliance. the satisfaction level of the customers. The customer expectation level
Spanning across an area of 55000 sq. ft. and a manpower of around 150 can be modelled as retailer's expectation level of the performance of the
employees, the state-of-the-art production facility is able to cater high suppliers in the evaluation process.
volumes of customer demands for home appliances. The growing Thus, the problem addressed in this study is to develop a supplier
market for home appliances has inspired the company to evolve with selection and allocation model based on customers’ sustainable ex-
newer ideas and product variants. When it comes to appliances, the pectations while considering the following aspects:
major focus of the company has always been to involve the voice of the
end-users in product development and designing, so that the products • Meeting company's target and customer's expectation: Suppliers must be
remain more user-centered and score well on market appeal. The major able to meet the customer expectations and goals of the company.
emphasis of the company (which is the focus of the current study) is These can be captured by evaluating the suppliers based on suitably
towards one of their home appliance- toaster. identified list of economic, environmental and social attributes.
The generic SC network of the company is shown in Fig. 1. The SC • Penalty and Bonus system: It is essential that the suppliers are se-
actors are manufacturer, distributors, retailers, suppliers, third party lected based on their sustainable score and the better performing
logistics providers (3PLPs) and customers. The raw material for man- suppliers are rewarded and low performing suppliers are penalized.
ufacturing the products is supplied by multiple suppliers at the • Cost vs Sustainable Allocation: Company must be able to optimise the
Fig. 1. Generic SC structure.
4
cost while allocating optimal amounts to the best performing sup- an optimisation model is developed in this phase, which integrates all
pliers so that the redesigned SC network brings economic stability the above aspects into a supplier selection model such that the total cost
and also enhances the sustainable performance of the company in and sustainable impact of the SC network (in terms of SVP) are both
the long run. optimised.
Phase 3: Weighted Goal programming approach is used to solve the
In order to incorporate the above aspects, an integrated model is bi-objective mathematical programming problem proposed in phase 2
developed in the study which aims to achieve a cost effective SC net- and in attaining trade-off between the objectives, significant to the
work by selecting the best performing sustainable suppliers and also decision environment.
optimally allocating purchase order among the selected suppliers. The A summary of the Methodology is pictorially described through
suppliers are evaluated through Z-TOPSIS process based on sustainable Fig. 2 provided below:
attribute which suitably represents the customers’ environmental and A detailed explanation of the three phases is provided in the fol-
social expectations. The evaluation scores which reflect the customers’ lowing subSections 4.1–4.3.
expectation level are then transformed as retailers’ expectations level
and are used as inputs in the optimisation model for selecting suppliers 4.1. Phase I: supplier evaluation process
and allocating orders based on the penalty and reward system. The
methodology adopted to develop the SC network model is discussed in 4.1.1. Identification of attributes
detail in the next section. In this study, customer sustainable (economic, environmental and
social) expectations have been translated into retailers’ expectations
4. Methodology based on the feedback that the retailers have received from the custo-
mers through a structured survey. Since, the link between manufacturer
The methodology adopted in the study for attaining the aim of the and customers are the intermediary retailers. Thus, on behalf of the
study involves the following three phases: manufacturing firm, the regional head has consulted retailers and taken
Phase 1: An exhaustive supplier evaluation process is performed their feedback into the decision making process. The group of DMs
which includes: included head of operations (having experience of 5–6 years), pro-
curement manager (having experience of 9–11 years), head of pur-
• Identification of vital sustainable attribute for the evaluation of chasing department (having experience of 7–10 years), marketing
manager (having experience of 5–8 years) and regional head (having
suppliers
• Computation of overall sustainable weights and socio-ecological experience of 8–12 years) as major DMs, is considered for evaluating
the suppliers’ performance. Eight attributes are extracted based on ex-
weights of suppliers using Z-TOPSIS
tensive literature survey and consensus with the DMs, as listed in
Phase 2: The evaluation of suppliers’ performance on the basis of Table 1.
significant sustainable attributes provides weights of importance of
suppliers. However, the selection of suppliers purely on the basis of 4.1.2. Fuzzy-TOPSIS method for assessment of suppliers
these assessment values is not what is desired by the company as sup- A Z-number is represented by an ordered pair of fuzzy numbers as
pliers performing upto the environmental and the social expectations of Z = (A˜ , B˜ ) where Ã signifies the fuzzy restriction and B̃ represents the
customers may demand higher cost. Further, along with the selection reliability of Ã [64]. Let A˜ = { x , µ Ã (x ) |x [0, 1]} and
decision the order allocation amounts based on the capacities of sup- B˜ = { x , µ B̃ (x ) |x [0, 1]} where µ Ã (x ) and µ B̃ (x ) are triangular fuzzy
plier and the holding capacity of retailers are also to be determined. It is membership functions.
also required that the amount of purchase order be optimally dis- Let there be ‘q’ DMs, q = 1, 2, …, Q, ‘k’ attributes, k = 1, 2, ,K
tributed among the selected suppliers based on their sustainability and ‘s’ suppliers, s = 1, 2, , S . The procedure of assessment of sup-
scores. To encourage the suppliers to perform better, the company also pliers using Z-TOPSIS approach is as follows:
wants to reward the better performing suppliers or impose penalty on Step 1: Evaluation of attributes by each DM with degree of con-
the poor performing suppliers them in terms of a monetary value. Thus fidence based on fuzzy linguistic terms defined in Tables 2 and 3.
Fig. 2. Proposed framework.
5
Table 1
Attributes for evaluation.
S. no. Attributes Description References
1 Product price (C1) Capability of selling the good quality product at reasonable price Luthra et al. [43]; Hashemi et al. [25]; Kannan
et al. [32]
2 Delivery efficiency (C2) Capability of delivering the product to the right location at the predetermined Luthra et al. [43]; Hashemi et al. [25]; Azadnia
time et al. [2]
3 Logistic & transportation Tendency of transporting the products as efficiently as possible Kannan et al. [32]; Luthra et al. [43]
efficiency (C3)
4 Technological capability (C4) Capability of adopting new technologies based on the current market demand Luthra et al. [43]; Hashemi et al. [25]; Azadnia
et al. [2]
5 Financial competence (C5) It refers to the financial aspects such as liquidity and profitability of the Büyüközkan & Çifçi [8]; Kuo and Lin [37]; Luthra
company et al. [43], Govindan et al. [22]
6 Flexibility (C6) Capability to handle variations in demand and lead time Luthra et al. [43]; Hashemi et al. [25]
7 Social conscience (C7) Suppliers’ responsibility towards employees’ rights protection, health and Luthra et al. [43]; Azadnia et al. [2]; Kannan et al.
safety. Level of commitment towards social enhancement of the SC. [32]
8 Green Competence and Energy Capability of altering the process and product which reduces effect on natural Kannan et al. [32]; Hashemi et al. [25]
Efficiency (C8) resources. Efficiently using energy as per the existing industry norms
Step 4: Transformation of weighted Z-number into Triangular Fuzzy

Table 2
Linguistic scale and TFN for evaluation of attributes.
Number (TFN) as follows:
Linguistic variable Triangular fuzzy number Z˜ = × A˜ = ( × l, × m, × n)
Very low (VL) (0,0,0.1) Step 5: Determination of attribute weights W

˜ = (w˜ 1, w˜ 2, ...,w˜ K ) by
Low (L) (0,0.1,0.3) aggregating the weighted Z-numbers of all DMs using the formula.
Medium low (ML) (0.1,0.3,0.5)
Medium (M) (0.3,0.5,0.7) 1 1
w˜ k = [w˜ k + w˜ k2 + ....+ w˜ kq]
Medium high (MH) (0.5,0.7,0.9) q
High (H) (0.7,0.9,1)
Very high (VH) (0.9,1,1) Step 6: Steps 1 to 5 are repeated to determine the aggregated
suppliers’ weight x̃sk for sth supplier with respect to kth attribute, using
the linguistic variables defined in Tables 3 and 4.
Step 7: Construction of fuzzy decision matrix D̃1 as follows:
Table 3
Linguistic scale and TFN for DMs’ reliability. x˜11 x˜12 … x˜1K
D˜ 1 = x˜21 x˜22 … x˜2K , s = 1, 2, …, S, k = 1, 2, …, K
Linguistic variable Triangular fuzzy number … … … …
x˜S1 x˜S1 … x˜SK
Strongly Unlikely (SU) (0,0,0.1)
Unlikely (U) (0,0.1,0.3) Where x˜sk =
1 1
[x˜sk + x˜sk2 + Q
+x˜sk ]
Q
Somewhat Unlikely (SWU) (0.1,0.3,0.5)
Neutral (N) (0.3,0.5,0.7) Step 8: Construction of normalized fuzzy decision matrix D̃2 as
Somewhat Likely (SWL) (0.5,0.7,0.9) follows:
Likely (L) (0.7,0.9,1)
Strongly Likely (SL) (0.9,1,1) D˜ 2 = [y˜sk ]SxK , s = 1, 2, ...,S, k = 1, 2, ...,K
Where
Step 2: Conversion of reliability assessment into crisp number using
lsk msk nsk
the given formula: y˜sk = , , and nk* = max nsk (Forbenefitattribute)
nk* nk* nk* s
xµ B˜ (x ) dx
= lk l l
µ B˜ (x ) dx y˜sk = , k , k and lk = min lsk (Forcostattribute)
nsk msk lsk s
Step 3: Determination of weighted Z-number for each DM as fol-
lows: Step 9: Computation of weighted normalized decision matrix
D̃3 using the following formula:
Z˜ = { x, µ A˜ (x ) |µ A˜ (x ) = µ A˜ (x ), x [0, 1] } D˜ 3 = [˜z sk]S × K , s = 1, 2, ...,S, k = 1, 2, ...,K
where z˜sk = y˜sk (.) w˜ k

Step 10: Computation of fuzzy positive ideal solution (FPIS) and
Table 4
fuzzy negative ideal solution (FNIS) for each alternative as follows:
Linguistic scale and TFN for evaluation of suppliers.
B * = (˜z1*, z˜*, zK* ) where z˜k* = (1, 1, 1), k = 1, 2, ...,K
2 ...,˜
Linguistic variable Triangular fuzzy number
Very poor (VP) (0,0,1) B = (˜z1 , z˜2 , ...,˜zK ) where z˜k = (0, 0, 0), k = 1, 2, ...,K
Poor (P) (0,1,3)
Medium poor (MP) (1,3,5) Step 11: Determination of distance of each alternative from FPIS
Fair (F) (3,5,7) and FNIS as follows:
Medium good (MG) (5,7,9)
K
Good (G) (7,9,10)
Very good (VG) (9,10,10) ds* = d z (˜z sk , z˜k*), s = 1, 2, ...,S
k=1
6
ds =
K
d z (˜z sk , z˜k ), s = 1, 2, ...,S
• α : penalty determination multiplier for performance value of sup-
1
plier being less than the retailer's expectation level
• market bonus multiplier for performance value of supplier being
k=1
β:
where dz (˜z sk , z˜k ) represents distance measurement between two fuzzy greater than the retailer's expectation level
numbers z̃ sk and z̃ k .
Step 12: Computation of closeness coefficient (CCs)of each alter-
• s : maximum number of suppliers that can be selected
max
native using equation: Decision variables

ds
CCs = , s = 1, 2, ...,S
• I : inventory of p product at r retailer in t time period
prt
th th th
• X : quantity flowing of p product from s supplier to r retailer in

ds + ds* th th th
psrt
Step 13: Rank the alternatives according to the closeness coeffi- th
t time period
cient. The best alternative is closest to the FPIS and farthest to the FNIS. • V : binary value taking value 1 if s supplier is selected, 0 otherwise
s
th
4.2. Bi-objective optimisation model

• : positive deviational variable measuring the increase in the socio-
+
s
ecological performance value of sth supplier for KSE, than the retailer
expectation level
The optimisation model is formulated as bi-objective programming
problem (BOP) under the following assumptions:
• s : negative deviational variable measuring the deficit in the socio-
ecological performance value of sth supplier for KSE, than the retailer
Assumptions expectation level
• All parameters are known and deterministic. Objectives

• Demands are divisible among suppliers and known with certainty. The cost objective given by Eq. (1) minimizes the total cost of the
• Multi-planning period is considered. system. The first term represents the procurement cost, the second term
• Single product having two different variants is considered. shows transportation cost, the third term corresponds to inventory
• Suppliers’ performance with respect to social and environmental holding cost and the last two terms shows the penalty and bonus
attribute below expectations is penalized and gets additional bonus amount.
if performance is above expectations.
• All suppliers have different capacities, lead time and procurement Min Total Cost = Min Z1 = Pcpsrt Xpsrt Vs
cost. p s r t
• Fixed capacities of suppliers and retailers are considered. + Tcpsrt Xpsrt Vs

• Quality of the product is same. p s r t
+ Hprt Iprt + 1 s Vs
The sets, parameters, decision variables, objectives and the con- p r t s
straints considered in the model are as follows: +

s Vs
Sets s
(1)
p ∈ PSet of products
Eq. (2) represents the objective of maximizing SVP which calculates
s ∈ S Set of suppliers
the total procurement amount while assigning the maximum order to
r ∈ R Set of retailers
suppliers with highest sustainable weights.
t ∈ T Set of time periods
k ∈ K Set of sustainable attributes where K = K Tr K SE Max Sustainable Value Purchase (SVP) = Max Z2
K Tr = {K1, K2, ...,K 6} is set of Traditional attributes and
= Xpsrt W soverall Vs
K SE = {K7, K 8} is set of Socio-ecological attributes p s r t (2)
Parameters Constraints
Eq. (3) restricts the minimum and maximum number of suppliers
• H : per unit inventory holding cost of p product at r retailer in t
prt
th th th
required.
time period
• Pc : procurement cost of p product supplied by s supplier to r
th th th 1 Vs smax
psrt
th
s (3)
retailer in t time period
• Tc : transportation cost per unit of p product transporting pro-
psrt
th th
th
th
Eqs. (4) and (5) are the inventory balancing equations of each re-
tailer for tth time period.
ducts from s supplier to r retailer in t time period
• D : demand of p product at r retailer in t time period
prt
th th th
• C : capacity of p product at the s supplier in t time period

1 th th th
Iprt = Ipr 0 + Xpsrt Vs Dprt p, r , t = 1
(4)
• C : capacity of p product at r retailer in t time period
pst s
2 th th th
• WI : Initial
prt
th th
Inventory of the p product at r retailer Iprt = Iprt + Xpsrt Vs Dprt p, r , t > 1
•
pr0 1
: performance evaluation scores of s supplier with respect to
overall th s (5)
s
all attributes obtained from Z-TOPSIS Eq. (6) limits the capacity of retailer in t time period. th
• E : retailer expectation with respect to socio-ecological perfor-

SE
2
mance of suppliers Iprt + Xpsrt Vs Cprt p, r , t
• Wsocio-ecological
SE
s
th
: performance evaluation score of s supplier with respect to the
attributes K = {K , K } obtained from Z-TOPSIS
SE
s (6)
7 8 Eq. (7) ensures that the demand of retailer is satisfied.
7
Table 5
Assessment of attributes by DMs.
K1 K2 K3 K4 K5 K6 K7 K8
(A,B) (A,B) (A,B) (A,B) (A,B) (A,B) (A,B) (A,B)
DM1 (VH,L) (H,N) (VH,L) (MH,N) (H,L) (M,N) (H,N) (H,L)

DM2 (VH,SL) (MH,L) (H,SL) (H,N) (VH,SL) (H,N) (VH,N) (VH,SL)
DM3 (H,L) (VH,L) (H,L) (H,L) (VH,L) (MH,L) (H,L) (VH,L)
DM4 (H,L) (H,L) (VH,L) (MH,L) (MH,L) (H,L) (VH,L) (H,L)
DM5 (VH,L) (MH,N) (H,SL) (M,N) (H,L) (H,N) (VH,N) (VH,SL)
Iprt + Xpsrt Vs Dprt p Step 2: In this step, negative and positive deviational variables ηi
r t s r t t r (7) and ρi are added in the objective functions and constraints as fol-
lows:
Eq. (8) restricts the capacity of each supplier in tth time period.
1 Zi (X ) + i i = Zi*, i = 1, 2
Xpsrt Vs Cpst s, r , t
p p (8)
Eq. (9) evaluates suppliers according to the expectations of retailers

with respect to the socio- environmental attributes. subject to gj (X ) + j j = bj ; j = 3, 4, ...,12
0
Vs (WsSE + s
+
s ) = E SEVs s (9)
i, i
Eq. (10) ensures that the performance evaluation scores of each of Step 3: The problem is first solved using rigid constraints only with
the selected supplier must be greater than the threshold. the objective to minimize the specific deviational variables as fol-
lows:
W soverall Vs 0.43*Vs s (10)
The non-negativity and binary restrictions are enforced by Eqs. (11) For Minimisation objective Z1 the deviational variableρ1is mini-
and (12) mised and for Maximisation objective Z2 the deviational variable η2is
+ minimised.
Iprt , Xpsrt , s , s 0 (11) For constraints if gj(X) ≤ bj then ρj is minimized if gj(X) ≥ bj then
ηjis minimized and if gj (X ) = bj then j + j
Vs {0, 1} (12)
The problem can be reformulated incorporating the optimal solu-
tion of the problem (P3) as follows:
4.3. Goal programming (GP) approach
Step 4: The following programming problem SOP3 is formulated for
Goal programming (GP), developed in early 60 s by Charnes and finding the compromised solution to the bi-objective problem de-
Cooper [11], is an efficient technique for dealing with multi-objective veloped in Section 4.2 (assuming all constraints are rigid).
optimization models having conflicting goals. The basic aim of GP is to SOP3
assign an aspiration level to each objective function. Afterwards, Minimize = 1 1 + 2 2
minimizes the weighted sum of undesirable deviances between actual Subject to
goals and aspiration levels. The most commonly used GP approach is (Z1/ Z1*) + 1 1 = 1
weighted goal programming (WGP). The objective of WGP is mini- (Z2/ Z2*) + 2 2 = 1
mizing the sum of deviations between goals and aspirations.
The steps for the GP approach are as follows: and constraints (3)-(12)where 1, 2 are normalised deviations from the
objectives and λi (i = 1,2) is the weight assigned to the ith objective.
Step 1: Solve the following Single Objective programming Problems The procedure given above helps in achieving the compromised
(SOPs) to obtain the aspiration levels for each objective: solution which helps in creating a trade-off between objectives.
SOP1:
Minimize Z1 5. Numerical illustration
Subject to constraints 3)–((12)
SOP2: The present study focuses on an example problem for a northern India
Maximize Z2 based home appliance manufacturing company. As discussed in the pro-
Subject to constraints 3)–((12) blem description (Section 3), the focus of the proposed study is to aid the
company in achieving their economic targets while satisfying the sus-
Let Z1*andZ2* be the optimal values of SOP1 and SOP2 respectively. tainability targets. In order to achieve this, the company needs to
Table 6
Attribute evaluation matrix with Z-numbers.
K1 K2 K3 K4 K5 K6 K7 K8
DM1 (0.84,0.93,0.93) (0.49,0.63,0.70) (0.84,0.93,0.93) (0.35,0.49,0.64) (0.65,0.84,0.93) (0.64,0.71,0.71) (0.49,0.63,0.70) (0.65,0.84,0.93)

DM2 (0.88,0.98,0.98) (0.47,0.65,0.84) (0.69,0.88,0.98) (0.49,0.63,0.70) (0.88,0.98,0.98) (0.49,0.63,0.70) (0.35,0.49,0.64) (0.88,0.98,0.98)
DM3 (0.65,0.84,0.93) (0.84,0.93,0.93) (0.65,0.84,0.81) (0.65,0.84,0.93) (0.84,0.93,0.93) (0.47,0.65,0.84) (0.65,0.84,0.93) (0.84,0.93,0.93)
DM4 (0.65,0.84,0.93) (0.65,0.84,0.93) (0.84,0.93,0.93) (0.47,0.65,0.84) (0.47,0.65,0.84) (0.65,0.84,0.93) (0.84,0.93,0.93) (0.65,0.84,0.93)
DM5 (0.84,0.93,0.93) (0.35,0.49,0.64) (0.69,0.88,0.98) (0.64,0.71,0.71) (0.65,0.84,0.93) (0.49,0.63,0.70) (0.35,0.49,0.64) (0.88,0.98,0.98)
8
Table 7
Fuzzy decision matrix for evaluation of suppliers w.r.t. sustainable attributes.
K1 K2 K3 K4 K5 K6 K7 K8
S1 (5.45,7.33,8.66) (3.03,4.80,6.58) (3.64,5.55,7.45) (2.87,4.46,5.91) (2.80,4.69,6.38) (3.05,4.64,5.90) (4.73,6.32,7.40) (3.25,5.15,7.06)

S2 (5.13,7.04,8.56) (4.08,5.78,7.20) (4.37,6.27,7.99) (3.34,5.05,6.55) (4.31,6.20,7.71) (4.89,6.67,7.88) (4.41,6.02,7.31) (4.80,6.70,8.22)
S3 (5.11,7.01,8.55) (4.95,6.55,7.79) (5.56,7.29,8.65) (5.40,7.08,8.38) (5.65,7.60,8.95) (4.71,6.33,7.60) (4.88,6.67,8.08) (5.11,6.83,8.17)
S4 (5.97,7.89,9.24) (6.35,7.66,8.27) (6.36,8.09,9.24) (6.30,7.89,8.78) (6.70,8.22,9.15) (6.13,7.36,7.93) (5.85,7.73,9.04) (7.09,8.64,9.43)
S5 (3.23,5.13,7.04) (2.11,3.70,5.29) (2.21,4.00,5.79) (3.52,5.16,6.79) (2.43,4.31,6.20) (2.29,4.06,5.83) (2.67,4.27,5.86) (2.22,4.03,5.85)
understand which suppliers to work with and who are able to meet their Table 9
expectation level in terms of socio-environmental aspects. In order to de- Final sustainable performance values of suppliers.
monstrate the validity the proposed model, a small data set of two variants Suppliers S1 S2 S3 S4 S5
marked as M1 and M2 of toasters is taken for four planning horizons. The
demand of 2 distributors catering to a total of 10 retail stores is considered. ds 2.10 2.30 2.53 2.66 2.13
For procuring the models, there is a pool of five suppliers (S = 5) who are d*
s 2.75 2.55 2.33 2.17 2.83
0.43 0.47 0.52 0.55 0.43
capable enough to supply as per the requirement of both the variants. The CCs = W soverall
manufacturer aims to evaluate the suppliers based on customers’ socio-
environmental expectations. The customer's expectation level is translated
in terms of retailer's expectation level of the supplier's performance and Similarly the aggregated weights of all suppliers are obtained and
based on that understanding a total of eight evaluation attributes (K = 8) the following fuzzy decision matrix D˜ 1 = [x˜sk ]5X 8 as shown in Table 7 is
are considered. The six traditional attributes and two socio-ecological at- constructed.
tributes are listed in Section 4.1.1. Next, normalized decision matrix D˜ 2 = [y˜sk ]5X 8 is constructed (pro-
Following Z-TOPSIS methodology given in Section 4.1.2, firstly the vided in Table A in the appendix).
decision opinions have been collected from five DMs (Q = 5) in terms For example for supplier S1 with respect to attribute K1 which a cost
of the linguistic variables and then fuzzy linguistic assessment scale attribute, l1 = min ls1 = 3.23and ỹ11 = (3.23/8.66, 3.23/7.33, 3.23/5.45) =
s
provided in Tables 2–4 in Section 4.1.2, is used to express them into Z- (0.37, 0.44, 0.59)
numbers. For example, the evaluation of all DMs for all attributes is Further, the weighted normalized matrix D˜ 3 = [˜z sk]5X 8is computed
provided by Table 5. In this table, the notation ‘VH’ in (VH, L) re- as shown in Table 8.
presents the importance of attribute K7 for DM4 is very high and value For example z˜11 = y˜11 . w˜ 1 = (0.37, 0.44, 0.59). (0.77, 0.90, 0.94) =
‘L’ signifies that the reliability of information provided by DM1 is likely. (0.29, 0.40, 0.56)
Similarly, the other values have been interpreted. Next, the FPIS and FNIS for each supplier are calculated as:
Next, the importance of attribute K7 given by DM4 is expressed in Z-
number as: Z˜ = (A˜ , B˜ ) = ((0.9, 1, 1), (0.7, 0.9, 1)) B * = [(1, 1, 1)1, (1, 1, 1)2 , ...,(1, 1, 1)5] and B
Further, the reliability of DM is converted into precise value and we = [(0, 0, 0)1, (0, 0, 0)2 , ...,(0, 0, 0)5]
get the following weighted Z-number: Z˜ = (A˜ , ) = ((0.9, 1, 1); 0.87)
The weighted Z-number is transformed into fuzzy number as: Subsequently, the distance of each supplier from FPIS and FNIS is
Z̃ = ( 0.87 × 0.9, 0.87 × 1, 0.87 × 1) = (0.84, 0.93, 0.93) calculated and the closeness coefficient is calculated as shown in
Repeating the process for each judgement, the weighted Z-numbers Table 9.
of all DMs with respect to attributes are obtained as shown Table 6. Similarly the evaluation scores of suppliers with respect to the social
The attribute weights obtained for all the DMs are aggregated using and environmental attribute set KSEare obtained which are as follows:
the average function and we get the following attribute weights: WsSE = (0.45, 0.39, 0.40, 0.40, 0.45) .
Next, in order to select the suppliers based on the above evaluation
˜ = (w˜ 1, w˜ 2, ...,w˜ 8)
W scores, the optimisation model developed in Section 4.2 needs to be
(0.77, 0.90, 0.94), (0.56, 0.71, 0.80), (0.74, 0.89, 0.93), solved. For this, the values of the parameters of the mathematical model
need to be defined as per the following information given by the company.
(0.44, 0.59, 0.72),
= The penalty determination multiplier if performance evaluation of
(0.70, 0.85, 0.92), (0.64, 0.76, 0.80), (0.35, 0.49, 0.64), supplier is less than the retailer expectation is taken as 3. While the
(0.74, 0.89, 0.95) market bonus multiplier if performance evaluation of supplier is greater
than the retailer expectation is taken as 2.
Table 8
Weighted normalized matrix for evaluation of suppliers w.r.t. sustainable attributes.
K1 K2 K3 K4 K5 K6 K7 K8
S1 (0.29,0.40,0.56) (0.20,0.41,0.64) (0.29,0.54,0.75) (0.14,0.30,0.49) (0.21,0.43,0.64) (0.25,0.44,0.59) (0.28,0.47,0.63) (0.23,0.38,0.65)

S2 (0.29,0.41,0.59) (0.28,0.49,0.70) (0.35,0.61,0.80) (0.17,0.34,0.54) (0.33,0.57,0.78) (0.40,0.64,0.79) (0.26,0.45,0.62) (0.20,0.30,0.44)
S3 (0.29,0.42,0.59) (0.33,0.56,0.76) (0.45,0.71,0.87) (0.27,0.48,0.69) (0.43,0.70,0.90) (0.38,0.61,0.76) (0.29,0.50,0.69) (0.20,0.29,0.41)
S4 (0.27,0.37,0.51) (0.43,0.65,0.80) (0.51,0.78,0.93) (0.31,0.53,0.72) (0.51,0.76,0.92) (0.50,0.71,0.80) (0.35,0.58,0.77) (0.17,0.23,0.30)
S5 (0.35,0.57,0.94) (0.14,0.32,0.51) (0.18,0.39,0.58) (0.17,0.35,0.56) (0.19,0.40,0.63) (0.19,0.39,0.59) (0.16,0.32,0.50) (0.28,0.49,0.95)
9
Table 10
Per unit inventory holding cost and demand of product for retailers.
Time period Holding cost (in INR)cost Demand (in units)
R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R1 R2 R3 R4 R5 R6 R7 R8 R9 R10
M1 t1 4.8 4.5 5 4.7 4.9 4.8 4.9 5 4.6 4.7 32 24 24 48 40 28 32 36 40 32

t2 4.7 4.4 4.9 4.6 5 4.8 5 4.9 4.7 4.6 40 32 32 44 48 36 40 40 48 36
t3 4.8 4.5 5 4.7 4.9 4.8 4.9 5 4.6 4.7 32 24 24 48 40 28 32 36 40 32
t4 4.7 4.4 4.9 4.6 5 4.8 5 4.9 4.7 4.6 40 32 32 44 48 36 40 40 48 36
M2 t1 4.6 4.7 4.5 4.8 4.9 4.7 4.8 4.9 4.7 4.8 28 32 28 28 40 36 32 44 40 28
t2 4.7 4.8 4.7 4.8 5 4.8 4.7 4.9 4.8 4.9 40 36 32 36 52 48 40 48 56 40
t3 4.6 4.7 4.5 4.8 4.9 4.7 4.8 4.9 4.7 4.8 28 32 28 28 40 36 32 44 40 28
t4 4.7 4.8 4.7 4.8 5 4.8 4.7 4.9 4.8 4.9 40 36 32 36 52 48 40 48 56 40
In this study, customer expectations has been translated into re- technique explained in Section 4.3 to obtain a trade-off between the
tailers’ expectations (ESE) with respect to socio-ecological performance contradictory goals of minimizing the total cost and maximising the
of suppliers. Customer satisfaction level has been identified based on SVP. To obtain the trade-off solution to the bi-objective problem, first
the survey. The survey was designed to complement the measures of the aspiration levels for both objectives are to be determined. For this,
feedback from customers on socio-ecological aspects. Retailers are re- initially two Single Objective Problems (SOPs) are solved over the same
quested to provide questionnaires to individual customers who visited set of constraints using LINGO 11.0: SOP1 with the “Total cost” ob-
to them in order to get the true picture of their satisfaction level. Thus, jective and SOP2 with “SVP” objective. The optimal objective values of
retailers’ expectation scores with respect to the social and environ- SOP1 and SOP2 are the highlighted values shown in Table 12. The
mental attribute setKSEare taken as:E SE = 0.40 . value of SVP obtained while minimising the total cost objective in SOP1
For each time period, the procurement amount to be determined by and the total cost incurred when maximising SVP objective in SOP2 are
the model must not exceed the demand of the retailers and it also also provided.
contributes towards the cost of the network. The demand of 10 retailers Using the aspirational values attained from the above table, first the
and the inventory holding costs are provided in Table 10. Also, the deviational variables are introduced in the two objectives and the soft
procurement amount ordered to each supplier must not exceed his ca- constraints. For the demand restrictions and SVP objective function,
pacity and the inventory to be stocked at each retail store must not under-achievement needs to be penalized. On the other hand, in case of
exceed the capacity of the retailer. Table 11 yields the capacities of capacity constraints and cost objective the over-achievement needs to
suppliers to deliver and the capacities of retailers for storing inventory. be penalized. The bi-objective problem is transformed to a single ob-
Further, the procurement cost of the model M1 and model M2 with cost jective problem (SOP3) of minimising the achievement function defined
per unit of transporting the finished goods from supplier to retail stores in terms of the unwanted normalized deviational variables. The
is given in Table B (in appendix). equivalent weighting structure by assigning the preferential weights of
The data considered here helps in solving the proposed SC network. all the unwanted deviations equal to one is being employed after dis-
The next section presents the results and provides its practical im- cussion with DMs. The optimal solution of SOP3 is provided in Table 13
plications. which is the trade-off solution to the BOP developed in Section 4.2.
Table C presents the inventory procured by the retail store in all time
6. Result discussion and implications periods.
6.1. Result 6.2. Result discussion
The Z-TOPSIS methodology discussed in Section 4.1 generates as- The aim of this research is to design an integrated framework for
sessment scores of the suppliers based on their performance which are selection of suppliers as to incorporate the customer sustainable ex-
further utilized in the bi-objective optimization model developed in pectations in the SC network of an Indian electronic manufacturer. The
Section 4.2. The mathematical model is further solved using WGP manufacturer wants to meet the sustainable expectation of the
Table 11
Capacities of the retailers and suppliers.
Time period Capacity of retailers Capacity of suppliers
R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 S1 S2 S3 S4 S5
M1 t1 40 32 28 56 44 36 36 40 44 36 140 152 160 128 120

t2 48 36 36 48 60 36 44 44 52 40 152 160 160 136 128
t3 40 32 28 52 44 36 36 40 44 36 140 152 160 128 120
t4 48 36 36 48 52 40 44 44 52 40 152 160 160 136 128
M2 t1 36 40 32 40 48 40 36 52 48 36 156 132 128 164 168
t2 44 40 36 40 56 52 44 52 60 44 160 140 140 176 180
t3 36 40 32 40 48 40 36 52 48 32 156 132 128 164 168
t4 44 40 36 40 60 52 44 52 60 44 160 140 140 176 180
10
Table 12
Optimal solutions of SOP1 and SOP2.
Optimisation problem Objective Total cost SVP Suppliers Sustainable weights of Socio-ecological weights of Procurement amount Deviation from retailers’ socio-environmental
(INR) selected selected suppliers selected suppliers expectation level
V.S. Yadavalli, et al.
SOP1 Minimisation of Total 4840663 1311.95 S1 S2 S3 S4 0.43 0.47 0.52 0.55 0.44 0.39 0.40 0.40 294 299 2493 103 + 0.05 - 0.01 + 0.01 + 0.01
cost
SOP2 Maximisation of SVP 5092599 1578.372 S3 S4 0.52 0.55 0.40 0.40 1110 1753 + 0.01 + 0.01
11
Table 13
Optimal solution of SOP3.
Optimisation problem Total cost (INR) SVP Suppliers selected Sustainable weights of selected Socio-ecological weights of selected Procurement amount Deviation from retailers’ socio-environmental expectation
suppliers suppliers level
SOP3 492363 1421.36 S1 S3 S4 0.43 0.52 0.55 0.44 0.40 0.40 290 1497 1402 + 0.05 + 0.01 + 0.01
Operations Research Perspectives 6 (2019) 100113
Fig. 3. Z-TOPSIS weights of suppliers.

Fig. 4. Pareto frontier.
customers, so as to work with suppliers with who are flexible enough suppliers purely on the basis of fulfilling customers’ expectations then it
towards the changing customer requirements. In order to incorporate will have to bear financial loss. However it can be observed that in both
the above aspects, the suppliers are evaluated through Z-TOPSIS pro- scenarios, suppliers selected have positive deviation from retailer's ex-
cess based on sustainable attribute which suitably represents the cus- pectation level of socio-environmental performance so no suppliers will
tomers’ environmental and social expectations in the first phase. Z- be penalised and would be rewarded. Overall, it can be stated that two
TOPSIS (steps given in Section 4.1) generates evaluation scores of the individual objectives are in conflict with each other in the attainment of
suppliers which reflect the customers’ expectation level are then the aspired values of the objectives.
transformed as retailers’ expectations level and are used as inputs in the This justifies the need for an effective solution technique such as
optimisation model for selecting suppliers and allocating orders based WGP utilised in the study to attain a compromised solution as per the
on the penalty and reward system. DMs’ opinions. The WGP model helps in attaining a compromised so-
To elaborate upon the evaluation process, in the first phase, we use lution with cost of INR 492363 and SVP value of 1421.36. Suppliers S1,
Z-TOPSIS generate the weights of the suppliers. As demonstrated S3 and S4 are selected.
through the Fig. 3, it is difficult to say which supplier is performing Thus, results clearly validate the effectiveness of proposed model by
better by simply looking at the scores obtained using Z-TOPSIS. Clearly, selecting best performing supplier while efficiently distributing pro-
the rankings are sensitive to the metric considered, and the final se- ducts among the best suppliers so that total cost of the system is opti-
lection based purely on the aggregated ranking may not be desirable. mized. The model has to compromise the cost for achieving the sus-
Hence, the mathematical model is proposed for the final selection based tainable value of purchase while fulfilling the customers’ expectations.
on certain SC constraints so that the overall aim of the DM is satisfied.
To capture the trade-off between the objectives, WGP approach is 6.3. Sensitivity analysis
utilized. Following the steps of the SOPs are solved individually.
Table 12 provides the details of the cost and SVP objectives along with The objective of the multi-objective optimization programming is to
the aspiration values. It can be seen from Table 12 that there exists find the most efficient solution which best satisfies all the objectives
conflict between the two objectives. The optimal solution of SOP1 is simultaneously. In the above problem, the aim is to find a “trade-off”
INR 4840663 in which suppliers S1, S2, S3 and S4 are selected with Z- solution which would give the best values of the cost and SVP objective
TOPSIS weights as 0.43, 0.47, 0.52 and 0.55 respectively with the SVP functions that are acceptable to the DM.
value obtained as 1312.116. Maximum number of suppliers is selected Therefore it is important to present a range of the trade-off solutions
with supplier 5 not selected Comparing with the high optimal SVP to the DMs to choose from, and understand the variation in the values
value of 1578.372 obtained after solving SOP2 shows that the first of both the objectives as the weights of importance the objectives are
model has clearly compromised on the sustainable performance of the changed, justifying the application of WGP. This can be used as quan-
suppliers in the selection decision. Similarly, the difference in the va- titative indicators for highlighting the efforts of the company towards
lues of the cost of the network obtained in SOP2 and SOP1, which is inclusion of socio-economic parameter into the decision making pro-
approximately INR 251936 shows that if the company wants to select cess. Based on the company's strategic vision, the DMs can choose the
Table 14
Sensitivity analysis.
Scenario Objectives Weights Objective values Suppliers selected Procurement amount Inventory
1 Minimisation of Total cost 0.5 INR 4972480 S1 S3 S4 2796 196

Maximisation of SVP 0.5 1541.435
2 Minimisation of Total cost 0.1 INR 5137873 S1 S4 2888 409
3 Minimisation of Total cost 0.9 INR 4857991 S1 S2 S3 S4 2792 200
12
most satisfactory solution from among the few efficient solutions pre- that the procurement amount is calculated as per the performance of
sented in Table 14. the supplier. The results also validate the claim that the model se-
It can be seen from the table that on assigning varying weights to lects the best performing suppliers and are also benefitted in terms
the objectives, the values of the cost and SVP vary. Although in scenario of getting higher procurement order. This is another way of giving
1 and 4, the same suppliers are selected, but the variations in the cost financial incentives to better performing suppliers.
objective are caused due the fluctuation in inventory and procurement
amount. In case of scenario 2, when we are giving more preference to 7. Conclusion
the SVP objective, the model yields a performance score of 1592.15
from the selection of suppliers S1 and S4. The focus of this scenario is This study contributes to the relevant area of research in the field of
more towards the selection of those suppliers who perform high on the sustainable supplier selection and order allocation problem by devel-
sustainable credentials, rather then selection those which will give least oping an integrated optimization model for selection and evaluation of
cost. The cost obtained in this scenario is INR 5137873, which clearly sustainable suppliers, reflection of customers’ needs and green ex-
demonstrates that the model has compromised on the cost profit to pectations and optimising firm's economic performance. The key fea-
select the suppliers who are high on sustainable performance. Similarly, tures of the integrated model are as follows: 1) selection of suppliers
it can be seen in the scenario 3, that when more preference to cost based on traditional as well as customers’ socio-ecological expectations
objective the selection is occurring in such a manner that least cost is transformed in terms of retailers’ expectation using fuzzy-TOPSIS
obtained, while satisfying the demand. Although more number of methodology with Z-numbers, 2) Incorporation of penalty-reward
suppliers are selected in this scenario yet the overall SVP is lower than system into the cost objective to award or penalize suppliers according
in the other scenarios. As can be seen from the above table that is to their socio-ecological performance, 3) allocation of appropriate order
difficult to say which solution is superior over the other (non-domi- quantity to best performing suppliers, 4) obtaining a trade-off between
nated solution), we plot all these and join then together in curve known two conflicting objectives of minimizing the SC cost and maximizing
as efficient frontier or pareto-optimal front. The Pareto optimal the SVP. The bi-objective mathematical model is solved using WGP to
curve of the problem is illustrated in Fig. 4. get the efficient solution. The model is then illustrated using a data set
All the solutions lying on the efficient frontier are efficient or pareto of an electronic organization to validate the results. The results clearly
optimal solutions to the problem. As a decision maker, it becomes easy relate the effectiveness of model. Outcomes of the proposed model can
to read the all possible alternative optimal solutions through a gra- be summarized as:
phical representation and select the one that suits the goals of the or-
ganization. • The main decision of the management concerned with the colla-
borating with sustainable suppliers, determining the optimum pur-
6.4. Implications chasing allocations while minimising the SC cost, is effectively
realised through development of an optimisation model.
The implications drawn from the result discussion lead to the • The penalty-reward system incorporated in the optimisation model
achievement of the research objectives of the study. These can aid is effective in providing incentives to better performing suppliers
managers in making crucial decisions as discussed below: and scope of improvement for other suppliers.
• This study helps the DMs in optimizing the cost of the system while The limitations of the current study are explained here which fur-
keeping customer expectation in consideration. The results show ther help in reinforcing the SC network design. First, SC network in this
that penalty and reward value incorporated in the cost objective paper depends upon the known demands. But, variations in demand
system is effective in enhancing the sustainable performance of the directly impact the performance of the organization which can be fur-
suppliers. Hence, penalising or rewarding suppliers based on their ther analysed. Secondly, the model developed in this model is specific
performance can prove to be an effective strategy for DMs. for an electronic sector. However, the model can be applied to other
• The objective of maximizing SVP in the mathematical model ensures industries also.
Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.orp.2019.100113.
Appendix
Table A, Table B, Table C.
Table A
Normalized matrix.
K1 K2 K3 K4 K5 K6 K7 K8
S1 (0.37,0.44,0.59) (0.37,0.58,0.79) (0.39,0.60,0.81) (0.33,0.51,0.67) (0.31,0.51,0.70) (0.38, 0.58, 0.74) (0.52,0.70,0.82) (0.31,0.43,0.68)
S2 (0.38,0.46,0.63) (0.49,0.70,0.87) (0.47,0.68,0.86) (0.38,0.57,0.75) (0.47,0.68,0.84) (0.62,0.84,0.99) (0.49,0.67,0.81) (0.27,0.33,0.46)
S3 (0.38,0.46,0.63) (0.60,0.79,0.94) (0.60,0.79,0.94) (0.61,0.81,0.95) (0.62,0.83,0.98) (0.59,0.80,0.96) (0.54,0.74,0.89) (0.27,0.32,0.43)
S4 (0.35,0.41,0.54) (0.77,0.93,1.00) (0.69,0.88,1.00) (0.72,0.90,1.00) (0.73,0.90,1.00) (0.77,0.93,1.00) (0.65,0.85,1.00) (0.24,0.26,0.31)
S5 (0.46,0.63,1.00) (0.25,0.45,0.64) (0.24,0.43,0.63) (0.40,0.59,0.77) (0.27,0.47,0.68) (0.29,0.51,0.74) (0.30,0.47,0.65) (0.38,0.55,1.00)
Attribute Weight (0.77,0.90,0.94) (0.56,0.71,0.80) (0.74,0.89,0.93) (0.44,0.59,0.72) (0.70,0.85,0.92) (0.64,0.76,0.80) (0.54,0.68,0.77) (0.74,0.89,0.95)
13
Table B
Procurement and transportation cost of products from suppliers to retailers.
Retailers Supplier Procurement cost Transportation cost
M1 M2 M1 M2
t1 t2 t3 t4 t1 t2 t3 t4 t1 t2 t3 t4 t1 t2 t3 t4
R1 S1 1760 1765 1760 1765 1605 1610 1605 1610 4.9 4.8 4.9 4.8 4.99 4.9 4.99 4.9
S2 1750 1755 1750 1755 1600 1620 1600 1620 5 4.9 5 4.9 5 4.9 5 4.9
S3 1860 1865 1860 1865 1760 1770 1760 1770 4.2 4.3 4.2 4.3 4.1 4.3 4.1 4.3
S4 1800 1810 1800 1810 1699 1700 1699 1700 4.8 4.9 4.8 4.9 4.89 4.92 4.89 4.92
S5 1897 1899 1897 1899 1799 1800 1799 1800 4.3 4.4 4.3 4.4 4.15 4.2 4.15 4.2
R2 S1 1750 1760 1750 1760 1699 1700 1699 1700 4.89 4.95 4.89 4.95 4.96 4.99 4.96 4.99
S2 1720 1730 1720 1730 1675 1680 1675 1680 4.99 5 4.99 5 5.01 5 5.01 5
S3 1830 1850 1830 1850 1760 1780 1760 1780 4.22 4.25 4.22 4.25 4.11 4.15 4.11 4.15
S4 1790 1799 1790 1799 1720 1750 1720 1750 4.88 4.9 4.88 4.9 4.95 4.99 4.95 4.99
S5 1860 1870 1860 1870 1788 1799 1788 1799 4.31 4.4 4.31 4.4 4.17 4.2 4.17 4.2
R3 S1 1799 1791 1799 1791 1615 1620 1615 1620 4.98 5.01 4.98 5.01 4.99 5 4.99 5
S2 1790 1799 1790 1799 1610 1605 1610 1605 5 4.99 5 4.99 5 4.99 5 4.99
S3 1880 1875 1880 1875 1770 1775 1770 1775 4.15 4.2 4.15 4.2 4.13 4.2 4.13 4.2
S4 1815 1810 1815 1810 1658 1660 1658 1660 4.87 4.93 4.87 4.93 4.98 5.01 4.98 5.01
S5 1887 1890 1887 1890 1787 1790 1787 1790 4.23 4.27 4.23 4.27 4.19 4.25 4.19 4.25
R4 S1 1810 1805 1810 1805 1710 1720 1710 1720 4.97 4.99 4.97 4.99 4.99 5 4.99 5
S2 1800 1810 1800 1810 1699 1700 1699 1700 5 5.02 5 5.02 5 5.01 5 5.01
S3 1867 1870 1867 1870 1775 1770 1775 1770 4.21 4.25 4.21 4.25 4.11 4.15 4.11 4.15
S4 1830 1835 1830 1835 1720 1725 1720 1725 4.92 4.95 4.92 4.95 4.98 5.01 4.98 5.01
S5 1880 1889 1880 1889 1780 1785 1780 1785 4.28 4.35 4.28 4.35 4.21 4.25 4.21 4.25
R5 S1 1810 1815 1810 1815 1700 1699 1700 1699 4.99 5.01 4.99 5.01 4.99 5 4.99 5
S2 1789 1790 1789 1790 1699 1705 1699 1705 5 4.99 5 4.99 5 4.99 5 4.99
S3 1860 1865 1860 1865 1758 1760 1758 1760 4.22 4.25 4.22 4.25 4.05 4.12 4.05 4.12
S4 1830 1840 1830 1840 1720 1725 1720 1725 4.98 4.99 4.98 4.99 4.99 5.01 4.99 5.01
S5 1878 1880 1878 1880 1760 1755 1760 1755 4.24 4.29 4.24 4.29 4.17 4.2 4.17 4.2
R6 S1 1805 1805 1805 1805 1690 1695 1690 1695 4.97 4.99 4.97 4.99 4.98 5 4.98 5
S2 1787 1790 1787 1790 1676 1680 1676 1680 5 5 5 5 5 5.02 5 5.02
S3 1857 1857 1857 1857 1778 1780 1778 1780 4.19 4.25 4.19 4.25 4.16 4.2 4.16 4.2
S4 1827 1830 1827 1830 1716 1719 1716 1719 4.96 4.99 4.96 4.99 4.95 4.99 4.95 4.99
S5 1869 1869 1869 1869 1780 1780 1780 1780 4.25 4.3 4.25 4.3 4.19 4.25 4.19 4.25
R7 S1 1799 1799 1799 1799 1750 1750 1750 1750 5 5.01 5 5.01 4.99 4.99 4.99 4.99
S2 1789 1790 1789 1790 1720 1720 1720 1720 5 4.99 5 4.99 5 5 5 5
S3 1870 1875 1870 1875 1778 1778 1778 1778 4.16 4.2 4.16 4.2 4.17 4.2 4.17 4.2
S4 1799 1800 1799 1800 1760 1760 1760 1760 4.98 5 4.98 5 4.95 4.99 4.95 4.99
S5 1879 1880 1879 1880 1784 1785 1784 1785 4.22 4.25 4.22 4.25 4.2 4.25 4.2 4.25
R8 S1 1789 1790 1789 1790 1696 1699 1696 1699 4.99 4.99 4.99 4.99 4.99 4.99 4.99 4.99
S2 1765 1770 1765 1770 1689 1689 1689 1689 5 5.01 5 5.01 5 5 5 5
S3 1873 1875 1873 1875 1772 1772 1772 1772 4.19 4.2 4.19 4.2 4.21 4.25 4.21 4.25
S4 1799 1800 1799 1800 1767 1770 1767 1770 4.97 4.99 4.97 4.99 4.97 4.99 4.97 4.99
S5 1882 1885 1882 1885 1786 1789 1786 1789 4.24 4.29 4.24 4.29 4.26 4.29 4.26 4.29
R9 S1 1789 1789 1789 1789 1722 1725 1722 1725 4.98 4.99 4.98 4.99 4.99 4.99 4.99 4.99
S2 1764 1769 1764 1769 1705 1710 1705 1710 5 5 5 5 5 5.01 5 5.01
S3 1875 1875 1875 1875 1775 1775 1775 1775 4.21 4.25 4.21 4.25 4.16 4.2 4.16 4.2
S4 1850 1850 1850 1850 1730 1730 1730 1730 4.9 4.99 4.9 4.99 4.98 4.99 4.98 4.99
S5 1888 1889 1888 1889 1785 1785 1785 1785 4.27 4.29 4.27 4.29 4.22 4.29 4.22 4.29
R10 S1 1822 1825 1822 1825 1720 1720 1720 1720 4.99 4.99 4.99 4.99 5 5 5 5
S2 1790 1790 1790 1790 1700 1700 1700 1700 4.99 5 4.99 5 5 5 5 5
S3 1878 1879 1878 1879 1783 1785 1783 1785 4.2 4.25 4.2 4.25 4.14 4.2 4.14 4.2
S4 1845 1845 1845 1845 1730 1730 1730 1730 4.98 4.99 4.98 4.99 4.99 4.99 4.99 4.99
S5 1890 1890 1890 1890 1788 1789 1788 1789 4.3 4.35 4.3 4.35 4.26 4.3 4.26 4.3
Table C
Inventory at retail stores for all time periods.
Product variants R1 R2 R3 R4 R5 R6 R7 R8 R9 R10
I1 4 13 9 11 9 13 9 9 9 5
I2 13 13 2 17 0 9 16 0 13 17
14
Conflict of interest
The authors have no conflict of interest.
References Prod 2013;56:164–72.

[28] Inderfurth K, Clemens J. Supply chain coordination by risk sharing contracts under
random production yield and deterministic demand. OR Spectrum
[1] Amindoust A, Ahmed S, Saghafinia A, Bahreininejad A. Sustainable supplier selec- 2014;36(2):525–56.
tion: a ranking model based on fuzzy inference system. Appl Soft Comput [29] Jadidi OMID, Zolfaghari S, Cavalieri S. A new normalized goal programming model
2012;12(6):1668–77. for multi-objective problems: a case of supplier selection and order allocation. Int J
[2] Azadnia AH, Saman MZM, Wong KY. Sustainable supplier selection and order lot- Prod Econ 2014;148:158–65.
sizing: an integrated multi-objective decision-making process. Int J Prod Res [30] Kannan D, Khodaverdi R, Olfat L, Jafarian A, Diabat A. Integrated fuzzy multi
2015;53(2):383–408. criteria decision making method and multi-objective programming approach for
[3] Bai C, Kusi-Sarpong S, Badri Ahmadi H, Sarkis J. Social sustainable supplier eva- supplier selection and order allocation in a green supply chain. J Cleaner Prod
luation and selection: a group decision-support approach. Int J Prod Res 2019:1–22. 2013;47:355–67.
[4] Bai C, Sarkis J, Wei X, Koh L. Evaluating ecological sustainable performance mea- [31] Kannan D, de Sousa Jabbour ABL, Jabbour CJC. Selecting green suppliers based on
sures for supply chain management. Supply Chain Manag Int J 2012;17(1):78–92. GSCM practices: using fuzzy TOPSIS applied to a Brazilian electronics company. Eur
[5] Bai C, Sarkis J. Integrating sustainability into supplier selection with grey system J Oper Res 2014;233(2):432–47.
and rough set methodologies. Int J Prod Econ 2010;124(1):252–64. [32] Kannan D, Govindan K, Rajendran S. Fuzzy Axiomatic Design approach based green
[6] Bai C, Sarkis J. Integrating sustainability into supplier selection: a grey-based supplier selection: a case study from Singapore. J Cleaner Prod 2015;96:194–208.
TOPSIS analysis. Tech Econ Develop Econ 2018;24(6):2202–24. [33] Kar AK. A hybrid group decision support system for supplier selection using analytic
[7] Bhutta KS, Huq F. Supplier selection problem: a comparison of the total cost of hierarchy process, fuzzy set theory and neural network. J Comp Sci 2015;6:23–33.
ownership and analytic hierarchy process approaches. Supply Chain Manag Int J [34] Kara SS. Supplier selection with an integrated methodology in unknown environ-
2002;7(3):126–35. ment. Expert Syst Appl 2011;38(3):2133–9.
[8] Büyüközkan G, Çifçi G. Evaluation of the green supply chain management practices: [35] Keshavarz Ghorabaee M, Amiri M, Zavadskas EK, Antucheviciene J. Supplier eva-
a fuzzy ANP approach. Prod Plan Control 2012;23(6):405–18. luation and selection in fuzzy environments: a review of MADM approaches. Econ
[9] Büyüközkan G, Göçer F. Application of a new combined intuitionistic fuzzy MCDM Res-Ekonomska istraživanja 2017;30(1):1073–118.
approach based on axiomatic design methodology for the supplier selection pro- [36] Kumar Kar A, K. Pani A. Exploring the importance of different supplier selection
blem. Appl Soft Comput 2017;52:1222–38. criteria. Manag Res Rev 2014;37(1):89–105.
[10] Chai J, Liu JN, Ngai EW. Application of decision-making techniques in supplier [37] Kuo RJ, Lin YJ. Supplier selection using analytic network process and data envel-
selection: a systematic review of literature. Expert Syst Appl 2013;40(10):3872–85. opment analysis. Int J Prod Res 2012;50(11):2852–63.
[11] Charnes, A., & Cooper, W.W. (1961). Management models and industrial applications [38] Kuo RJ, Wang YC, Tien FC. Integration of artificial neural network and MADA
of linear programming (No. 339.23 C4). methods for green supplier selection. J Cleaner Prod 2010;18(12):1161–70.
[12] Choudhary D, Shankar R. A goal programming model for joint decision making of [39] Laari S, Töyli J, Solakivi T, Ojala L. Firm performance and customer-driven green
inventory lot-size, supplier selection and carrier selection. Comp Indust Eng supply chain management. J Cleaner Prod 2016;112:1960–70.
2014;71:1–9. [40] Li L, Zabinsky ZB. Incorporating uncertainty into a supplier selection problem. Int J
[13] Coskun S, Ozgur L, Polat O, Gungor A. A model proposal for green supply chain Prod Econ 2011;134(2):344–56.
network design based on consumer segmentation. J Cleaner Prod 2016;110:149–57. [41] Li S, Ragu-Nathan B, Ragu-Nathan TS, Rao SS. The impact of supply chain man-
[14] Dobos I, Vörösmarty G. Green supplier selection and evaluation using DEA-type agement practices on competitive advantage and organizational performance.
composite indicators. Int J Prod Econ 2014;157:273–8. Omega 2006;34(2):107–24.
[15] Dweiri F, Kumar S, Khan SA, Jain V. Designing an integrated AHP based decision [42] Lin PC, Huang YH. The influence factors on choice behavior regarding green pro-
support system for supplier selection in automotive industry. Expert Syst Appl ducts based on the theory of consumption values. J Cleaner Prod 2012;22(1):11–8.
2016;62:273–83. [43] Luthra S, Govindan K, Kannan D, Mangla SK, Garg CP. An integrated framework for
[16] Fallahpour A, Olugu EU, Musa SN, Khezrimotlagh D, Wong KY. An integrated model sustainable supplier selection and evaluation in supply chains. J Cleaner Prod
for green supplier selection under fuzzy environment: application of data envel- 2017;140:1686–98.
opment analysis and genetic programming approach. Neur Comp Appl [44] Mathiyazhagan K, Govindan K, NoorulHaq A, Geng Y. An ISM approach for the
2016;27(3):707–25. barrier analysis in implementing green supply chain management. J Cleaner Prod
[17] Gopal PRC, Thakkar J. Sustainable supply chain practices: an empirical investiga- 2013;47:283–97.
tion on Indian automobile industry. Prod Plan Control 2016;27(1):49–64. [45] Mathur PP, Shah J. Supply chain contracts with capacity investment decision: two-
[18] Govindan K, Khodaverdi R, Jafarian A. A fuzzy multi criteria approach for mea- way penalties for coordination. Int J Prod Econ 2008;114(1):56–70.
suring sustainability performance of a supplier based on triple bottom line ap- [46] Moghaddam KS. Fuzzy multi-objective model for supplier selection and order al-
proach. J Cleaner Prod 2013;47:345–54. location in reverse logistics systems under supply and demand uncertainty. Expert
[19] Govindan K, Rajendran S, Sarkis J, Murugesan P. Multi criteria decision making Syst Appl 2015;42(15–16):6237–54.
approaches for green supplier evaluation and selection: a literature review. J [47] Mohammed A, Harris I, Kannan G. A hybrid MCDM-FMOO approach for sustainable
Cleaner Prod 2015;98:66–83. supplier selection and order allocation. Int J Prod Econ 2019.
[20] Govindan K, Sivakumar R. Green supplier selection and order allocation in a low- [48] Nazari-Shirkouhi S, Shakouri H, Javadi B, Keramati A. Supplier selection and order
carbon paper industry: integrated multi-criteria heterogeneous decision-making allocation problem using a two-phase fuzzy multi-objective linear programming.
and multi-objective linear programming approaches. Annal Oper Res Appl Math Modell 2013;37(22):9308–23.
2016;238(1–2):243–76. [49] Ordoobadi SM. Development of a supplier selection model using fuzzy logic. Supply
[21] Govindan K, Agarwal V, Darbari JD, Jha PC. An integrated decision making model Chain Manag Int J 2009;14(4):314–27.
for the selection of sustainable forward and reverse logistic providers. Annal Oper [50] Orji IJ, Wei S. An innovative integration of fuzzy-logic and systems dynamics in
Res 2017:1–44. sustainable supplier selection: a case on manufacturing industry. Comp Indust Eng
[22] Govindan K, Darbari JD, Agarwal V, Jha PC. Fuzzy multi-objective approach for 2015;88:1–12.
optimal selection of suppliers and transportation decisions in an eco-efficient closed [51] Rezaei J, Nispeling T, Sarkis J, Tavasszy L. A supplier selection life cycle approach
loop supply chain network. J Cleaner Prod 2017;165:1598–619. integrating traditional and environmental criteria using the best worst method. J
[23] Güçdemir H, Selim H. Customer centric production planning and control in job Cleaner Prod 2016;135:577–88.
shops: a simulation optimization approach. J Manuf Syst 2017;43:100–16. [52] Rostamzadeh R, Govindan K, Esmaeili A, Sabaghi M. Application of fuzzy VIKOR for
[24] Hammami R, Temponi C, Frein Y. A scenario-based stochastic model for supplier evaluation of green supply chain management practices. Ecol Indic
selection in global context with multiple buyers, currency fluctuation uncertainties, 2015;49:188–203.
and price discounts. Eur J Oper Res 2014;233(1):159–70. [53] Rouyendegh BD, Saputro TE. Supplier selection using integrated fuzzy TOPSIS and
[25] Hashemi SH, Karimi A, Tavana M. An integrated green supplier selection approach MCGP: a case study. Proced-Soc Behav Sci 2014;116:3957–70.
with analytic network process and improved Grey relational analysis. Int J Prod [54] Sarkis J, Talluri S. A model for strategic supplier selection. J Supply Chain Manag
Econ 2015;159:178–91. 2002;38(1):18–28.
[26] Ho W, Xu X, Dey PK. Multi-criteria decision making approaches for supplier eva- [55] Scott J, Ho W, Dey PK, Talluri S. A decision support system for supplier selection
luation and selection: a literature review. Eur J Oper Res 2010;202(1):16–24. and order allocation in stochastic, multi-stakeholder and multi-criteria environ-
[27] Hsu CW, Kuo TC, Chen SH, Hu AH. Using DEMATEL to develop a carbon man- ments. Int J Prod Econ 2015;166:226–37.
agement model of supplier selection in green supply chain management. J Cleaner [56] Selviaridis K, Norrman A. Performance-based contracting for advanced logistics
15
services: challenges in its adoption, design and management. Int J Phys Distrib Logist Prod Econ 2015;170:178–90.
Manag 2015;45(6):592–617. [62] Wisner JD, Tan KC, Leong GK. Principles of supply chain management: a balanced
[57] Shen L, Olfat L, Govindan K, Khodaverdi R, Diabat A. A fuzzy multi criteria ap- approach. Cengage Learning; 2014.
proach for evaluating green supplier's performance in green supply chain with [63] Yazdani M, Chatterjee P, Zavadskas EK, Zolfani SH. Integrated QFD-MCDM fra-
linguistic preferences. Resour Conserv Recycl 2013;74:170–9. mework for green supplier selection. J Cleaner Prod 2017;142:3728–40.
[58] Starbird SA. Penalties, rewards, and inspection: provisions for quality in supply [64] Zadeh LA. A note on Z-numbers. Inform Sci 2011;181(14):2923–32.
chain contracts. J Oper Res Soc 2001;52(1):109–15. [65] Zhang X, Xu Z. Soft computing based on maximizing consensus and fuzzy TOPSIS
[59] Tseng SC, Hung SW. A framework identifying the gaps between customers' ex- approach to interval-valued intuitionistic fuzzy group decision making. Appl Soft
pectations and their perceptions in green products. J Cleaner Prod 2013;59:174–84. Comput 2015;26:42–56.
[60] Vahidi F, Torabi SA, Ramezankhani MJ. Sustainable supplier selection and order [66] Zhu Q, Sarkis J, Lai KH. Institutional-based antecedents and performance outcomes
allocation under operational and disruption risks. J Cleaner Prod of internal and external green supply chain management practices. J Purchas Supply
2018;174:1351–65. Manag 2013;19(2):106–17.
[61] Wang W, Zhang Y, Zhang K, Bai T, Shang J. Reward–penalty mechanism for closed- [67] Aissaoui N, Haouri M, Hanssini E. Supplier Selection and order lot sizing modelling:
loop supply chains under responsibility-sharing and different power structures. Int J A review. Comput Oper Res 2007;34(12):3510–40.
16

Costumer Satisfaction 4

Uploaded by

Copyright:

Available Formats

Costumer Satisfaction 4

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Costumer Satisfaction 4

Uploaded by

Copyright:

Available Formats

Operations Research Perspectives 6 (2019) 100113

Contents lists available at ScienceDirect

Operations Research Perspectives

An integrated optimization model for selection of sustainable suppliers T

ARTICLE INFO ABSTRACT

Fig. 1. Generic SC structure.

Fig. 2. Proposed framework.

Step 4: Transformation of weighted Z-number into Triangular Fuzzy

Linguistic variable Triangular fuzzy number Z˜ = × A˜ = ( × l, × m, × n)

Very low (VL) (0,0,0.1) Step 5: Determination of attribute weights W

where z˜sk = y˜sk (.) w˜ k

native using equation: Decision variables

• X : quantity flowing of p product from s supplier to r retailer in

4.2. Bi-objective optimisation model

• All parameters are known and deterministic. Objectives

• Fixed capacities of suppliers and retailers are considered. + Tcpsrt Xpsrt Vs

straints considered in the model are as follows: +

• C : capacity of p product at the s supplier in t time period

• E : retailer expectation with respect to socio-ecological perfor-

DM1 (VH,L) (H,N) (VH,L) (MH,N) (H,L) (M,N) (H,N) (H,L)

Eq. (9) evaluates suppliers according to the expectations of retailers

DM1 (0.84,0.93,0.93) (0.49,0.63,0.70) (0.84,0.93,0.93) (0.35,0.49,0.64) (0.65,0.84,0.93) (0.64,0.71,0.71) (0.49,0.63,0.70) (0.65,0.84,0.93)

S1 (5.45,7.33,8.66) (3.03,4.80,6.58) (3.64,5.55,7.45) (2.87,4.46,5.91) (2.80,4.69,6.38) (3.05,4.64,5.90) (4.73,6.32,7.40) (3.25,5.15,7.06)

S1 (0.29,0.40,0.56) (0.20,0.41,0.64) (0.29,0.54,0.75) (0.14,0.30,0.49) (0.21,0.43,0.64) (0.25,0.44,0.59) (0.28,0.47,0.63) (0.23,0.38,0.65)

M1 t1 4.8 4.5 5 4.7 4.9 4.8 4.9 5 4.6 4.7 32 24 24 48 40 28 32 36 40 32

6.1. Result 6.2. Result discussion

M1 t1 40 32 28 56 44 36 36 40 44 36 140 152 160 128 120

Fig. 3. Z-TOPSIS weights of suppliers.

1 Minimisation of Total cost 0.5 INR 4972480 S1 S3 S4 2796 196

Table A, Table B, Table C.

The authors have no conflict of interest.

References Prod 2013;56:164–72.

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.