Journal Search Engine
Search Advanced Search Adode Reader(link)
Download PDF Export Citaion korean bibliography PMC previewer
ISSN : 1598-7248 (Print)
ISSN : 2234-6473 (Online)
Industrial Engineering & Management Systems Vol.19 No.2 pp.308-322

Optimizing Manufacturing Supplier Selection and Order Allocation Processes using Fuzzy-QFD and Goal Programming: A Case Study in Indonesian Chemical Industry

Suhartini*, Anindya Rachma Dwicahyani, Suparno, Hari Supriyanto
Department of Industrial Engineering, Adhi Tama Institute of Technology Surabaya (ITATS), Surabaya, Indonesia
Department of Industrial Engineering, Institut Teknologi Sepuluh Nopember (ITS), Surabaya, Indonesia
Corresponding Author, E-mail:
March 1, 2019 July 8, 2019 January 16, 2020


This empirical study aims to implement an optimal methodology of a supplier selection process by integrating Fuzzy Quality Function Deployment (FQFD) and Goal Programming (GP). We apply the method to a company engaged in chemical industry, under multi item, multi supplier, and multi period environments. Problems encountered in this study are how to select suppliers and determine the optimal order allocation in accordance with company requirements. We use FQFD and GP to evaluate and allocate orders to suppliers. Four objectives are incorporated, including maximization of good products, on time delivery, and supplier scoring, along with minimization of material purchasing cost. We seek to find an optimal solution considering several constraints, such as material requirement and cost estimation, supplier and warehouse capacity, and safety stock. Our results provide insights that the most and the least important technical responses in chemical supplier selection are “supplier experience” and “response to customer order”, respectively. Whereas, the most important criteria are “product quality”, “product price”, and “delivery performance”, while the least important criteria are “return & repair service” and “documents”. The result also shows that the proposed method can achieve better solution compared to the past decision, in term of cost and supplier scoring.



    The emerging era of global markets along with the fourth industrial revolution have resulted in an increasingly tight manufacturing competition. In order to preserve their existence, companies are required to improve its performance continuously. The performance of manufacturing companies is generally measured through several criteria such as production process, product quality, ability to fulfil customer satisfaction, up to distribution system. One of the factors that improve the production flow is the performance of suppliers. This is related to the function of suppliers itself, which is providing raw materials as main component in the production process.

    The role of suppliers in supply chain management is very important (Bevilacqua et al., 2006;Ghorabaee et al., 2017). The process of supplier selection must be done by the right method to help management in making optimal procurement decision. Within the company, between 50%-90% of the tasks in production department are decision regarding the procurement and operational strategies. Improvements in related areas can be done systematically and transparently to select, evaluate, and allocate orders to suppliers.

    In many developing countries, such as Indonesia, manufacturing companies often struggle with designing a process that help them to overcome problems related to raw material procurement processes. Many Indonesian companies still carry out the process of evaluating and selecting suppliers traditionally that only based on the past experiences and intuitions without applying a scientific approach. Therefore, due to its poor planning, various problems arise resulting in inefficient processes and increasing costs. In fact, various Multi-Criteria Decision Making (MCDM) approaches can actually be applied in the supplier selection process, such as Analytical Hierarchy Process (AHP), Analytical Network Process (ANP), Quality Function Deployment (QFD), TOPSIS, etc. The implementation of these scientific methods needs to be carried out by manufacturing companies to solve the problems related to raw materials procurement processes.

    Among all of the MCDM methods, many researchers still consider QFD as one of the most powerful tools to overcome various kind of decision-making problems. Latest development of QFD in the field of supply chain management has been conducted by Rozar et al. (2019) in which they developed a QFD method to improve GSCM sustainability performance of Malaysia’s Manufacturing SMEs. From the study, they found out that social factor is the most important parameter of a GSCM improvement. Whereas, economy, operational, and environment are other parameters that affecting GSCM improvement of manufacturing SMEs. Nikolaeva (2018) integrated QFD along with lean production supply chain in case of the process of providing medical care to patients in a service sector. In the study, he focused on waste and cycle time reduction, along with analysis of the decreasing amount of refusals from hospitalization.

    In addition to supplier evaluation and selection, another challenge is how to optimally allocate orders to suppliers. In term of materials order allocation, manufacturing companies are also facing various constraints confining the system to make strategic decisions. These constraints include material requirements, product prices, maximum purchasing capacity, warehouse capacity, and safety stock. Therefore, it requires an approach that is able to provide optimal solutions based on the condition. Goal Programming (GP) is one multi-objective optimization approach that is able to find a compromise solution by considering the targets and constraints of the company. Through out the application of GP, several objectives can be accommodated within the criteria to minimize deviation between those.

    Research on supplier selection have been carried out by many researchers using various approaches. Table 1 compares several prior studies in the related field of research. Ghodsypour and O’Brien (1998) used AHP and Linear Programming (LP) with the objective function of maximizing Total Purchase Value (TPV). Research in the related field was also carried out by Wang et al. (2004) in which they used AHP and GP to maximize Total Purchase Value (TPV) and minimize Total Purchase Cost (TPC). Demirtas and Üstün (2005) used ANP and GP with the objective function of minimizing total material purchasing cost, minimizing defective rate, and maximizing total purchase value. They considered constraints related to capacity and demand. Furthermore, Bevilacqua et al. (2006) used Fuzzy-Quality Function Deployment (FQFD) to evaluate suppliers. The application of fuzzy set theory gives flexibility to accommodate uncertainty due to the vague information and subjective preference elements. Fuzzy set theory also allows the measurement of vague aspirations from decision makers and its participation in the decision-making process.

    Kilic (2013) developed a supplier evaluation process by integrating fuzzy TOPSIS and Mixed-Integer Linear Programming (MILP). Evaluation was done for multiitem / multi-supplier environment. The study examined a case of water filter company. They consider several criteria, including quality, cost, delivery time, geographical location, and references. Shad et al. (2014) developed a methodology of supplier selection using QFD, Fuzzy AHP, and Linear Physical Programming (LPP) under uncertainties. Tavakoli and Pasha (2015) integrated FQFD with Fuzzy Linear Goal Programming (FLGP) to select supplier. However, the FLGP approach was used to obtain the weight of criteria. Whereas, they used FQFD approach to rank the suppliers. The problem of order allocation was not investigated in the paper. Case study was undertaken in PETROCHEMICAL Co. in Iran.

    Singh (2016) developed a supplier evaluation model with Fuzzy TOPSIS, MILP, and GP approaches. They aimed to minimize inventory and transportation costs, and to maximize procurement value (PV). Several constraints were considered, including demand condition, supplier capacity, budget, and delivery lead-time. In the computational experiment, they incorporated five criteria of supplier, including reliability, quality, on time delivery, consistency, and price. In medical sector, development of QFD and fuzzy set theory was conducted by Karsak and Dursun (2015) and Dursun and Karsak (2015). Erginel and Gecer (2016) developed a Fuzzy Multi Objective Linear Programming (FMOLP) to select supplier the medical sector. Lima-Junior and Carpinetti (2016) proposed a FQFD method in supplier selection to include the type of item and the type of supply chain in determining requirements and criteria.

    One most recent study in the related field was done by Abdel-Basset et al. (2018). In the study, they developed a framework of supplier selection by using a neutrosophic set theory and AHP-QFD. A case study on a pharmaceutical manufacturing company in Egypt was presented and seven criteria of supplier evaluation were used. The interesting point of the study is that, they used neutrosophic set theory instead of fuzzy set theory. They explained that neutrosophic set is more efficient and flexible rather than fuzzy set. The neutrosophic set was also applied by Van et al. (2018) to develop a new QFD approach to evaluate and select supplier in the green supply chain management.

    Zaheri et al. (2018) developed a model of bi-level programming to solve supplier selection and lot sizing problem. In their study, three optimization methods were applied and compared, i.e. Kurus-Kuhn-Tucker (KKT), Particle Swarm Optimization (PSO), and Differential Evolution (DE). The developed algorithm supports supplier selection and determination of the optimal production and delivery policies that minimize the total costs. The model of Zaheri et al. (2018) does not use MCDM approach, but rather use a mathematical apporach. They use quality criteria to evaluate suppliers, which is the value of Cpk. The result showed that the model can incorporate the buyer and suppliers simultaneously.

    Another study was also done by Babbar and Amin (2018) in which they developed a supplier selection and order allocation model in the case of beverages industry. They applied fuzzy QFD along with Stochastic Mixed Integer Linear Programming (SMILP). They included the aspect of environmental concern as one criteria in the evaluation. Azadnia and Ghadimi (2018) developed a FAHPQFD and Fuzzy Mixed Integer Non Linear Programming (FMINLP) to solve problem of supplier selection and order allocation. In the study, they include 4 objectives including minimization of the annual purchasing cost, maximization of the economic sustainability, the environmental sustainability, as well as the social sustainability. The problem is subjected to several constraints including maximum demand, maximum delivery, maximum order capacity, and maximum quality rate of the delivered products. The model was applied in a food manufacturing company.

    With an effort to overcome problem related to supplier selection and order allocation in the scope of Indonesian chemical industry, we develop a decision making approach using FQFD and GP. FQFD method is a powerful method to capture a set of suitable attributes of evaluation criteria based on company requirement. In this case, due to the lack of company knowledge, the company is considered to be less aware of the right and appropriate criteria for evaluating suppliers. The company needs a methodology that can translate their requirements into more structured criteria. Therefore, FQFD is used as a basis for the supplier selection to measure the performance of supplier candidates. Along with FQFD, we use GP to allocate orders to suppliers. We propose seven technical responses as evaluation criteria. The determination of seven technical responses is based on the company requirements in the case of chemical industry. The results of FQFD, which are supplier scores, are then used as one objectives function in the GP model to allocate orders to suppliers along with minimization of costs and maximization of non-defective products and on-time deliveries. Several constraints are also incorporated in the proposed model, including material requirement, material cost estimation, maximum order size, maximum warehouse capacity, and safety stock. Determination of the objectives and constraints is based on the characteristics and requirements of the related company, which in this case is a chemical industry. We propose the GP model and consider it to be comprehensive to overcome problem of order allocation in the related industry.

    1.1 Problem Description: Case Study in Indonesian Chemical Industry

    The problem of this study is how to evaluate and select supplier as well as allocate order of each type of material, to the selected supplier, at each period of time according to company requirements. We accommodate fuzziness in decision-making process in the proposed model. The purposes are to select suppliers in accordance with company criteria and determine the optimal order allocation to the selected suppliers. The case taken in this study is a procurement process of Indonesian chemical industry. In the investigated system, the company has two options in purchasing raw materials, i.e. domestic and foreign (international) suppliers. If company orders raw materials to foreign suppliers, then orders will be delivered through the port of Tanjung Perak and then distributed to the company with the help of MKL (Shiploads Expedition Company) and PBM (loading and unloading company). The investigated process is depicted in Figure 1.


    This section explains the methodology and stages for solving problems that have been defined previously. The method used in this study is Fuzzy Quality Function Deployment (FQFD) and Goal Programming (GP). FQFD is applied to evaluate supplier candidates, whereas GP is used to determine the optimal order allocation for selected suppliers. The research methodology is given in Figure 2.

    The initial stage of this research is a preliminary study to carry out a thorough observation and analysis of the problems in the procurement section of Indonesian chemical industry. The next step is to identify the problem and the objectives of the research. Data collection begins by identifying company requirement as well as supplier information.

    2.1 Supplier Evaluation using Fuzzy QFD

    According to Tavakoli and Pasha (2015), QFD has been said as an efficient tool to translate customer requirements (WHATs) into appropriate technical characteristics (HOWs). The application of QFD is very common in various sectors of manufacturing industry. QFD is flexible, useful, and effective to provide an integrated planning framework in many areas including technology, academic, and improvement projects, business planning, as well as manufacturing strategic planning (Tidwell and Sutterfield, 2012). The applications of QFD are facing many challenges, one of them is the relationship between WHATs and HOWs which are often vague and imprecise (Bevilacqua et al., 2006). Researchers then developed a Fuzzy QFD (FQFD) that incorporated vagueness in the input variables. FQFD uses linguistic variables that expressed in fuzzy numbers for describing inputs.

    In this study, we apply the FQFD method to evaluate and select suppliers. In addition to the effectiveness and popularity of QFD in the area of manufacturing strategic planning, the method is applied because of its compatibility with the problem that is to be resolved in this study. Supplier evaluation criteria are determined based on company requirements, and then translated into several technical responses. In this case, due to the lack of company knowledge, the company is considered to be less aware of the right and appropriate criteria for evaluating suppliers. Therefore, the company needs a methodology that can translate their requirements into more structured criteria. These criteria are then weighted based on their relative importance toward others.

    The first step of this study is to form a team of expert from 12 heads of department in a Indonesian chemical industry. The expert team includes three departments of production, department of maintenance, quality control, logistic, marketing, laboratory, transportation, HR, finance, and purchasing. The questionnaire method is done in the following steps.

    2.1.1 Identify the Company Requirements (WHATs)

    In this step, preliminary questionnaires of supplier evaluation criteria were deployed and given to the expert team. The result of the preliminary study shows that out of the 25 initial criteria, there are 12 criteria that considered important and will subsequently be used to evaluate suppliers. Table 2 gives details of the chosen criteria.

    2.1.2 Define the Linguistic Values and Corresponding Fuzzy Number

    After the company requirements have been identified, the next step is to define the linguistic values and its corresponding fuzzy number that used for evaluation. In this study, we used a triangular fuzzy number with five linguistic values, e.g. very insignificant (“Very Low” / (VL)), not important (“Low” / (L)), important (“Medium” / (M)), important (“High” / (H)), and very important (“Very High” / (VH)). Each of these linguistic values is plotted on a fuzzy set given in Table 3. The fuzzy set is denoted as {α, β, Γ}.

    2.1.3 Define the Weight of Importance of Criteria

    After the company requirements have been identified, then we define the weight of importance of each criteria based on an assessment through questionnaire. The assessment was carried out by 3 decision makers, including the head of HR department, head of transportation department, and head of purchasing department. The assessment was done in five linguistic values given in previous section. Subsequently, the next step is to aggregate the decision maker’s estimations regarding importance level of criteria that have been collected through questionnaire. We use the additive method to aggregate the weight of importance of each criteria as shown in Equation (1), with q denotes an index for company requirement (q = {1, 2, …, Q}).

    w q α ,     β ,   γ = 1 3 × { w q α +   w q β +   w q γ }   for q = { 1 , 2 , , Q }

    2.1.4 Identify the Technical Responses (HOWs)

    This section explains how company criteria (WHATs) are translated into technical responses. The company criteria in Table 1 are translated into several technical responses shown in Table 4. In this case, a criterion can be translated into more than one technical responses.

    Decision makers assessed the relationship between technical responses (HOWs) that have been identified in Table 4. The aim was to find out how changes in technical responses affect each other.

    2.1.5 Identify the Relation between Company Requirements and Technical Responses

    Decision makers also assessed the relationship between company requirements (WHATs) and technical responses (HOWs). The result is summarized in Table 4. The value is aggregated using additive method shown in Equation (2), with r denotes an index for company requirement (q = {1, 2, …, Q}), t denotes an index for technical responses (t = {1, 2, …, T}), and m denotes an index for decision maker (m = {1, 2, …, M}).

    r t q = 1 M × m   =   1 M r t q m for t = { 1 , 2 , , T } , q = { 1 , 2 , , Q }

    2.1.6 Identify the Weight of Importance for Technical Responses

    Decision makers assessed the weight of the technical response (how) in supplier selection using questionnaire method. The result is summarized in Table 5. We use the additive method to aggregate the weight of importance of each technical responses as shown in Equation (3).

    w t = 1 T q = 1 Q r t q × w q for  t = { 1 , 2 , , T }

    2.1.7 Develop the HOQ Matrix

    After collecting data, the next step is to construct a House of Quality (HoQ) diagram that describes the overall information needed, in order to develop and determine service quality in supplier evaluation.

    2.1.8 Supplier Ranking and Scoring

    The final step of the supplier evaluation is to score and rank the supplier candidate based on the existing technical responses. FSI and Score is calculated using Equation (4) and (5), respectively, with SRjt denotes rating of supplier j based on technical response t.

    F S I j = 1 T t = 1 T S R j t × w t   for  j = { 1 , 2 , , J }

    S c o r e j = F S I j α + 2 × F S I j β + F S I j γ 4 for  j = { 1 , 2 , , J }

    2.2 Goal Programming Model Development

    2.2.1 Objectives of the Model

    We seek to determine the optimal order allocation of material i, to supplier j, at period k, which denoted by Xijk. We include more than one objective using the method of goal programming. There are four objectives incorporated in the study, including:

    1. Maximize the good products indicated by percentage of non-defective products obtained from supplier j. Formulation of objective 1 (l = 1), denoted by Z1, is given in Equation (6) below.

      Z 1 =   j = 1 J k = 1 K r i j k   X i j k for   i = { 1 , 2 , , I }

    2. Minimize the total cost of purchase from supplier j with cost estimation of c. Formulation of objective 2 (l = 2), denoted by Z2, is given in Equation (7).

      Z 2 = j = 1 J k = 1 K ( c i j k ) X i j k for i = { 1 , 2 , , I }

    3. Maximize the on-time delivery from supplier j. Formulation of objective 3 (l = 3), denoted by Z3, is given in Equation (8) below.

      Z 3 = j = 1 J k = 1 K l i j k   X i j k for  i = { 1 , 2 , , I }

    4. Maximize the supplier scoring. Supplier score is the result of supplier evaluation using Fuzzy QFD. This objective was accommodated to consider qualitative factors in supplier selection. Qualitative factors include company requirements and expectations of what should be provided by the services of suppliers. This objective is also needed to ensure that the allocations given to suppliers are in accordance with the performance of each supplier. Formulation of objective 4 (l = 4), denoted by Z4, is given in Equation (9) below.

    Z 4 = j = 1 J k = 1 K b i j k   X i j k for   i = { 1 , 2 , , I }

    2.2.2 Minimum Deviation between Objectives

    The next step of goal programming is to minimize deviation between objectives. All of the objectives are compromised in the function of deviation variable. The goal is to minimize the deviation according to the objective that has been set as priority. The goal function is the total alternative performance that has been added to underestimate, reduced overestimate and must be equal to the constraints as given in Equation (10).

    Z i l + n i l p i l = g i l for i = { 1 , 2 , , I } , and l = { 1 , 2 , , L }

    Constraints are needed as consideration to support the decision making process. Constraints provide an option to choose an alternative that gives the same results in achieving the goal. In this case, several constraints are considered including material requirement, material cost estimation, maximum order size, warehouse capacity, and safety stock. These five constraints will be included in the constraint functions, so that each constraint function has deviation variables, in the form of positive and negative deviations. These constraints are formulated below.

    1. Material requirement/demand constraint

      • In order to ensure that the number of products ordered to suppliers is in accordance with company’s demands, we incorporate the number of material requirement which is obtained from MRP process. Given Dk denotes demand at period k, the formulation regarding material requirement constraint (g1) is given by Equation (11) below.

      g 1 i = j = 1 J X i j k + n 1 i p 1 i = D k for i = { 1 , 2 , I } , k = { 1 , 2 , , K }

    2. Material cost estimation/budget constraint

      • In this case, material prices must be in accordance with the material cost estimation. In other words, the total material purchasing cost must be lower or equal to company’s budget allocation. Given ce denotes material cost estimation and cij denotes material price of material i from supplier j, the formulation regarding maximum price constraint is given as

      g 2 i = j = 1 J c i j X i j k + n 2 i p 2 i c i j D k for i = { 1 , 2 , I } , k = { 1 , 2 , , K }

    3. Maximum order size/order capacity constraint

      • This constraint aims to ensure that the material ordered to each supplier does not exceed the maximum shipping capacity of suppliers. Given Qijk denotes maximum shipping capacity of material i, from supplier j, at period k, the function of the related constraint can be formulated as

      g 3 i = X i j k + n 3 i p 3 i Q i j k for i = { 1 , 2 , I } , k = { 1 , 2 , , K }

    4. Maximum warehouse capacity

      • As we consider a warehouse with limited capacity, we incorporate constraint related to this capacity. Given C denotes capacity of warehouse, we formulate the constraint as

      g 4 i = j = 1 J X i j k + S S k S S ( k 1 ) + n 4 i p 4 i C i for i = { 1 , 2 , I } , k = { 1 , 2 , , K }

    5. Safety stock

      • This constraint aims to ensure that buffer stocks are always available in the warehouse. Given vik is an amount of material-i set by company as a buffer stock at period-k, the safety stock constraint can be formulated as

      g 5 i = S S i k + n 5 i p 5 i = v i k for i = { 1 , 2 , I } , k = { 1 , 2 , , K }

    At this stage, the objective function is to minimize the deviation from all constraints. The objective function of deviation minimization is given in Equation (16).

    l = 1 L n l i + p l i for i = { 1 , 2 , , I }


    Case presented in this study is the procurement process of Liku Telaga Ltd., an Indonesian company that engaged in the chemical industry. Liku Telaga Ltd. is located in Gresik, one of the cities in East Java province, Indonesia. The company is an associate company of Lautan Luas Inc., which produces Sulphur acid and aluminium sulfate. In this case, we investigate two types of material, including Sulphur and Aluminium Hydroxide, with the period of six months.

    Company and supplier information are collected, including number of supplier candidate, supplier selling price, supplier capacity, warehouse capacity, amount of raw materials required for production process, and safety stock. In the case of Liku Telaga Ltd., there are four candidates of supplier for Sulphur including Standard Chemical Corp. Pte. Ltd., Lautan Luas Ltd., Yosomulgo Jajag Ltd., and Archindo Ltd. Whereas for Aluminium Hydroxide, there are four candidates of supplier, including Bisindo Kencana Ltd., Hindalco Industries Ltd., Sumitomo, and Chemindus Inc. The supplier data are given in Table 5, 6, and 7.

    The company data regarding material requirement for 6 periods, warehouse capacity, and safety stock are given in Table 8 and 9, respectively.

    3.1 Supplier Evaluation using FQFD

    The identification of company requirement is done by interviewing and deploying questionaire to 12 heads of department that are directly related to the company procurement process. Several information are collected, including company needs and requirements (see Table 2), technical responses (see Table 4), weight of importance of each criteria, relationship between company requirements and technical responses, and supplier assesments. Table 10 shows the weight of importance of each criteria. The values also represent the priority of company requirements in supplier selection.

    Table 11 summarizes the assessment result regarding the relationship between company requirements (WHATs) and technical responses (HOWs). The next step is to aggregate the weigth of importance of each technical response (HOWs). The result is given in Table 12.

    Finally, we calculate the value of FSI and score of each supplier, and determine the rank of each supplier. The supplier FSI, score, and rank are given in Table 13.

    3.2 Order Allocation using Goal Programming

    The order allocation problem is solved using the procedure of GP discussed in Section 2.2. Several constraints that we concerned include material requirements, material cost estimation, maximum order size, warehouse capacity, and safety stock based on the data in Tables 5, 8, and 9. The optimization is done by the help of Lingo Software. The optimal result of order allocation of material i, to supplier j, at period k for the related case study is given in Table 14.


    4.1 Optimal Order Allocation using Goal Programming

    The optimization result of goal programming, given in Table14, shows orders to suppliers at each period have fulfilled all of the company's material requirements. For Sulphur materials, the company consistently orders the most to Chemical Corp. Standard. Pte. Ltd. (see Figure 3). As for Aluminium Hydroxide, the company orders most to Bisindo Kencana Ltd. (see Figure 4).

    Figure 3 and 4 gives the optimal order allocation to each supplier for Sulphur and Aluminium Hydroxide, for period 1 to 6. Based on the results for those types of material, it is known that the largest quantity of material is not ordered to the supplier with highest score (based on the evaluation with QFD), but rather to the supplier with second highest score. The reason is that the proposed model not only considers the score of the supplier, but also considers several other objectives, including product quality, purchasing cost, and the accuracy of delivery. This finding must be considered by decision makers in selecting and allocating orders to suppliers. If the company chooses a supplier based solely on the results of the QFD evaluation, the solution obtained may not be the best one. In this case, integration between QFD and goal programming needs to be done in order to obtain a best solution by combining qualitative and quantitative assessment. Therefore, it is expected that the final solution is truly optimal that has incorporated various aspects of consideration.

    4.2 Analysis of the Results of Achieving Goals as Output of Goal Programming

    In this study, the proposed model has incorporated several objective including maximizing non-defective products, maximizing on time delivery, minimizing total purchasing cost and maximizing supplier scoring. Several constraints are also included in optimizing the model. The optimal value of each objective are given in Table 15.

    For Sulphur material, the objective function of maximizing good products (Z1) gives a value of 4,809,167 tons (see Table 15). The maximum good product that can be achieved is 98% from the total company order, hence we get that the defective rate is 2%. The objective function to minimize total purchasing price (Z2) is $513,720. From this point, we understand that the company achievement is 1% exceeding the desired target. The third objective function is to maximize the on time delivery (Z3), the model achievement on Z3 is 4,791.5 tons from the total order of 4,900 tons. This means that 98% of the shipments are delivered on time. The fourth objective function, maximize the supplier scoring (Z4), a maximum value of 1,229.24 tons is obtained. The result indicates the ability of selected suppliers to provide orders. Similar results are also obtained for Alumunium Hydroxide, with Z1 = 6,813.6 tons, Z2 = $ 2,137.5, Z3 = 6,763.6 tons, and Z4 = 1,779.34 tons, with defective rate of 2% and on time delivery rate of 97%.

    4.3 Analysis of the Integrated FQFD and Goal Programming Method

    Quality Function Deployment (QFD) is used to evaluate suppliers according to company requirements. From these requirements, critical criteria emerge. To meet these critical criteria, technical responses are needed. Decision makers assessed and evaluated supplier candidates based on technical responses that already determined by management. The assessment results in rank and score of suppliers which will be included in one of the objective functions of GP. We use GP to determine the optimal order allocation of each supplier. Of the four objective functions incorporated in the model, the results show the ability of each supplier, based on the technical response determined by the company.

    In this case study, Lautan Luas Inc. as a supplier for Sulphur material has the highest score of 200.8 with values of FSI = (109.8, 193.6, 306.9) and Hindalco Industries Ltd. as suppliers for aluminium hydroxide material having the highest score of 219 with FSI = (121.7, 211.6, 331.3). From the results of GP, order allocation indicates that the supplier with the highest score or good technical response ability is not necessarily getting the largest order quantity. This can be seen in the Standard Chemical Corp. Pte. Ltd. which obtained an order amounted to 3,646.7 tons, while Lautan Luas Inc. only obtained an order of 803.3 tons. Bisindo Kencana Ltd. obtained an order of 3,000 tons, while Hindalco Industries Ltd. only obtained an order of 1,114 tons.

    To clarify the achievements of using this method to the case study, we provide a table of comparison between the past decisions and the proposed one as summarized in Table 16.

    Formerly, the company only considered quality, price, and delivery performance as criteria to evaluate suppliers (see Table 16). The past decisions were to select Standard Chemical Corp. Pte. Ltd. and Hindalco Industries Ltd. as suppliers of Sulphur and aluminium hydroxide. Using the proposed FQFD and GP method, seven criteria were considered, including supplier experience, capacity fulfilment, certification system of quality, response to customer order, financial condition, communication, and geographical location. By applying the proposed method, all of the supplier candidates are selected with different amount of orders. As shown in Table 16, the proposed method provides savings on total cost of purchase of $17,702 for both Sulphur and aluminium hydroxide materials.


    The issue of supplier selection continues to be a concern for manufacturing industry, including the chemical industry. This study develops an integration model of Fuzzy Quality Function Deployment (FQFD) with goal programming to evaluate and allocate orders to suppliers. FQFD is built on the conditions and problems experienced by chemical companies in Indonesia. There are twelve company requirements (WHATs) and seven technical responses (HOWs) proposed to evaluate suppliers. Assessment was carried out with a questionnaire to management, related to the WHATs and HOWs criteria. In this study, the applied Goal Programming incorporated four objective functions, i.e. maximizing good products, maximizing on-time delivery, maximizing supplier scoring, and minimizing total cost of purchases. Several constraints are involved to find the optimal solutions, including material requirement, material cost estimation, maximum order size, warehouse capacity, and safety stock constraints. The decision variable is the optimal order allocation for each type of material, to each supplier, at each period of time.

    The case study was conducted at an aluminium sulfate manufacturer in Indonesia. By applying the proposed method, suppliers with the highest score are Lautan Luas Inc., as Sulphur supplier, with a score of 200.8, and Hindalco Industries Ltd., as aluminium hydroxide supplier, with a score of 219.0. The most important technical response is supplier experience, followed by capacity fulfillment, location, financial condition, certification system of quality, communication, and response to customer orders. The proposed method is more economical since it provides a saving on total cost of purchase of $17,702. In addition, the proposed method can also guarantee that the defectives rate can be minimized and on-time delivery can be maximized without ignoring the supplier performances.

    The GP result shows that the supplier with the second highest score actually gets the largest order, instead of the first highest scored supplier. The proposed method do not only consider supplier scores as its objective, but also total cost of purchase, delivery performance, and non-defective products. The result of this study emphasizes that companies should choose and allocate orders not only based on supplier scores, but also other considerations. Accommodation of various considerations will result in a better optimal solution.

    For future development, this research can be continued by applying not only FQFD, but also Fuzzy- Goal Programming. We suggest that it is better to consider weight or priority in the objective function to select and allocate order to suppliers. Another possible development is to further analyze the BCOR (Benefit, Cost, Opportunity and Risk) and model sensitivity. In addition, in the order allocation model, shipping time and shipping costs should also be considered.



    The raw material procurement process investigated in the study.


    Research methodology.


    Optimal order allocation to each supplier for sulphur.


    Optimal order allocation to each supplier for aluminium hydroxide.


    Comparison of this study to several prior studies of the related supplier selection and order allocation problem

    Details of company requirements (WHATs)

    Convertion of linguistic values to their corresponding fuzzy numbers

    Technical responses (HOWs)

    Data of supplier regarding material price and delivery capacity

    Delivery performance of suppliers

    Product quality of suppliers

    Material requirements of Liku Telaga Ltd. for 6 periods

    Warehouse capacity and safety stock of Liku Telaga Ltd.

    Weight of importance of each criteria (aggregated)

    Relationship between company requirement (WHATs) and technical responses (HOWs)

    Weight of importance of each technical responses (aggregated)

    Supplier FSI, score, and rank based on decision maker assessment

    Optimization result of order allocation using GP

    The optimal value of each objective

    Comparison of the past decision and the proposed method


    1. Abdel-Basset, M. , Manogaran, G. , Mohamed, M. , and Chilamkurti, N. (2018), Three-way decisions based on neutrosophic sets and AHP-QFD framework for supplier selection problem, Future Generation Computer Systems, 89, 19-30.
    2. Azadnia, A. H. and Ghadimi, P. (2018), An integrated approach of fuzzy quality function deployment and fuzzy multi-objective programming to sustainable supplier selection and order allocation, Journal of Optimization in Industrial Engineering, 11(1), 1-22.
    3. Babbar, C. and Amin, S. H. (2018), A multi-objective mathematical model integrating environmental concerns for supplier selection and order allocation based on fuzzy QFD in beverages industry, Expert Systems With Applications, 92, 27-38.
    4. Bevilacqua, M. , Ciarapica, F. E. , and Giacchetta, G. (2006), A fuzzy-QFD approach to supplier selection, Journal of Purchasing and Supply Management, 12(1), 14-27.
    5. Demirtas, E. A. and Üstün Ö., (2005), Analytic network process and goal programming approach for multi-period lot-sizing with supplier selection, Proceedings of the 35th International Conference on Computers and Industrial Engineering, 513-518.
    6. Demirtas, E. A. and Üstün Ö., (2008), An integrated multiobjective decision making process for supplier selection and order allocation, Omega, 36(1), 76-90.
    7. Dursun, M. and Karsak, E. E. (2015), Fuzzy Decision Approach Based on QFD and FWA for Selection of Medical Suppliers, Proceedings of the World Congress Engineering, II, London, U.K.
    8. Erginel, N. and Gecer, A. (2016), Fuzzy multi-objective decision model for calibration supplier selection problem, Computers & Industrial Engineering, 102, 166-174.
    9. Ghodsypour, S. H. and O’Brien, C. (1998), A decision support system for supplier selection using an integrated analytic hierarchy process and linear programming, International Journal of Production Economics, 56-57, 199-212.
    10. Ghorabaee, M. K. , Amiri, M. , Zavadskas, E. K. , and Antucheviciene, J. (2017), Supplier evaluation and selection in fuzzy environments: A review of MADM approaches, Economic Research-Ekonomska Istraživanja, 30(1), 1073-1118.
    11. Karsak, E. E. and Dursun, M. (2015), An integrated fuzzy MCDM approach for supplier evaluation and selection, Computers & Industrial Engineering, 82, 82-93.
    12. Kilic, H. S. (2013), An integrated approach for supplier selection in multi-item/multi-supplier environment, Applied Mathematical Modelling, 37(14-15), 7752-7763.
    13. Lima-Junior, F. R. and Carpinetti, L. R. (2016), A multicriteria approach based on fuzzy QFD for choosing criteria for supplier selection, Computers & Industrial Engineering, 101, 269-285.
    14. Nikolaeva, A. (2018), Lean production supply chain and QFD-analysis in the process of healthcare, International Journal of Supply Chain, 7(6), 528 - 535.
    15. Rozar, N. M. , Razik, M. A. , and Zakaria, M. N. (2019), Sustainability performance approach in Malaysia’s SMEs for improving green supply chain management (GSCM): An application of quality function deployment (QFD), International Journal of Supply Chain Management, 8(2), 993-998.
    16. Sajedinejad, A. and Chaharsooghi, S. K. (2018), Multi-criteria supplier selection decisions in supply chain networks: A multi-objective optimization approach, Industrial Engineering & Management Systems, 17(3), 392-406.
    17. Shad, Z. , Roghanian, E. , and Mojibian, F. (2014), Integration of QFD, AHP, and LPP methods in supplier development problems under uncertainty, Journal of Industrial Engineering International, 10(1), 1-9.
    18. Singh, A. (2016), A goal programming approach for supplier evaluation and demand allocation among suppliers, International Journal of Integrated Supply Management, 10(1), 38-62.
    19. Tavakoli, M. and Pasha, N. (2015), Integrating fuzzy quality function deployment and linear goal programming for supplier selection, Uncertain Supply Chain Management, 1-10.
    20. Tidwell, A. and Sutterfield, J. S. (2012), Supplier selection using QFD: A consumer products case study, International Journal of Quality & Reliability Management, 29(3), 284-294.
    21. Van, L. H. , Yu, V. F. , Dat, L. Q. , Dung, C. C. , Chou, S.-Y. , and Loc, N. V. (2018), New integrated quality function deployment approach based on interval neutrosophic set for green supplier evaluation and selection, Sustainability, 10(3), 1-13.
    22. Wang, G. , Huang, S. H. , and Dismukes, J. P. (2004), Product-driven supply chain selection using integrated multi-criteria decision-making methodology, International Journal of Production Economics, 91(1), 1-15.
    23. Zaheri, F. , Zandich, M. , Taghavifard, M. T. , and Najafi, E. (2018), A bi-level programming model to solve supplier selection and lot-sizing problem addressing quantity discounts and transportation cost, Industrial Engineering & Management Systems, 17(2), 267-280.