Documents
Resources
Learning Center
Upload
Plans & pricing Sign in
Sign Out

SUPPLIER SELECTION MODEL FOR PRODUCT DESIGN

VIEWS: 9 PAGES: 11

									20          International Journal of Electronic Business Management, Vol. 8, No. 1, pp. 20-30 (2010)


                         A SUPPLIER SELECTION MODEL
                        FOR PRODUCT DESIGN CHANGES
              Zhen-Hua Che1, Tzu-An Chiang2, Chuang Tu3 and Cheng-Jui Chiang1
                    1
                      Department of Industrial Engineering and Management
                            National Taipei University of Technology
                                      Taipei (106), Taiwan
                    2
                     Department of Commerce Automation and Management
                            National Pingtung Institute of Commerce
                                     Pingtung (900), Taiwan
                    3
                      Department of Industrial Engineering and Management
                                Chienkuo Technology University
                                    Changhua (500), Taiwan

                                                ABSTRACT
            Environmental pollution has significantly influenced ecology and has thus become a global
            issue. To satisfy diverse market demands and to upgrade product competitiveness,
            companies have prioritized the production of green products. In the initial stage of new
            product production, companies need to evaluate green component suppliers for those that
            meet environmental requirements in product design changes. By following the WEEE and
            RoHS directives of the EU, and by eliminating component suppliers with negative green
            indexes, this study constructed an optimal decision-making model for the selection of green
            component suppliers. The supplier selection was conducted by considering a quantity
            discount mechanism under the scenario of seeking minimum total cost. Lastly, this study
            obtained the solution for the model by applying Particle Swarm Optimization (PSO), and
            constructed a decision support system upon the decision-making process for the reference
            of decision makers.

            Keywords: Product Design Change, Supplier Selection, Quantity Discount, Particle Swarm
            Optimization

             1. INTRODUCTION                               the costs of the products are related to component
*                                                          procurement costs. Thus, this study suggested that
       Product production is often modified due to         when selecting product plans, it is critical to
clients’ demands. Thus, design change of products is       recognize the most suitable component supplier
a necessary activity for companies. Jonghoon and Lee       combination and quantity allocation.
[17] suggested that the model design of the products             In addition, environmental pollution is a serious
should be modified frequently, which would affect          issue of concern. Deans [5] indicated that industries
the subsequent manufacturing and production costs.         in the U.S. are promoting environmental protection,
Thus, evaluating product plans effectively at the          and environmental factors are significant to
initial stage of any change and selecting the              procurement policy and suppler selection. In 2003,
appropriate product plan are critical to reducing          the EU announced WEEE and RoHS directives,
overall production costs successfully. However, most       which had great impact on the electronics industry.
studies on the selection of product plans tend to target   After the official execution of these two directives,
strategies [19,4,3], and are seldom concerned with the     products in violation of the regulations would be
component supply of the product plans. Thus, the           denied entry into the EU nations. Leading companies
selected product plans may possibly fail due to            in the electronics industry have recognized the
inefficient production, as suppliers are critical to the   importance of these directives, and have proposed
success of the products. Dowlatshahi [6] suggested         strategies leading to the management of a green
that in order to facilitate new product development,       supply chain. Lai [20] suggested the green supply
companies must effectively integrate the suppliers’        chain framework and strategy, and indicated that
resources. Handfield [12] indicated that over 50% of       these strategies would affect all members in the
                                                           supply chain, including the brand suppliers, design
*                                                          companies, manufacturing companies, and materials
    Corresponding author: wind112233@yahoo.com.tw
Z. H. Che et al.: A Supplier Selection Model for Product Design Changes                                      21

suppliers. Che [2] indicated that only products                   2. LITERATURE REVIEW
meeting the WEEE and RoHS directives were
allowed to be exported to European countries under       2.1 Selection of Green Suppliers
the    environmental      protection    consciousness.          With the rise in environmental consciousness,
Manufacturers need a decision-making model to find       governments have forced companies to improve their
suppliers that could comply with WEEE and RoHS           environmental outcomes. In order to respond to green
derivatives, while maintaining low cost, short           issues, in the past decade, companies have begun to
production and delivery time, and high quality.          formulate green plans to organize their supply chains
      In addition, Shin & Benton [28] and Wang &         from the view of environmental efficacy [22,14].
Che [31] illustrated that quantity discount had been a   Noci [23] suggested that to manage environmental
key strategy in supplier selection. To meet actual       issues effectively, companies must integrate the
demands, the quantity discount should be introduced      relationship between customers and suppliers. In fact,
in decision making for supplier selection. Sirias and    they should incorporate relations with the production
Mehra [26] argued that normal manufacturers used         of green products and the suppliers in order to
discounts out of three motives: 1. to reach the best     achieve the following results: (1) reduction in the
price with respect to one consumer, or a group of        quantity of components with low environmental
similar consumers, 2. to reach partial price, with       performance, (2) effective control of the costs of
respect to a group of similar consumers, and 3. to       green products, (3) reduction in the time to respond to
affect a buyer’s order configuration. In general,        market demands.
discount in discussions can influence a buyer’s order          Sarkis [25] proposed a strategic decision-
configuration, or a manufacturer‘s specific discounts    making framework for managers, which emphasized
in an attempt, other than increasing the orders of       the elements of green supply chain management and
downstream manufacturers. Therefore, this study          how these elements are turned into the basic factors
applied quantity discount to the supplier selection      of a decision-making framework. Five strategies were
process for product design change in order to be more    included: product life cycle strategy, environmental
complete and realistic. Tsai [29] proposed that price    strategies for the organization, organizational efficacy
and quantity formed a nonlinear model due to             demands, and plans of a green supply chain.
quantity discount variables but under such conditions,   Humphreys [16] indicated that traditional supplier
it was difficult to find a global optimal solution.      selection only targeted quality and flexibility.
Hence, this study tried to solve this problem by         However, due to stresses from environmental issues,
employing the PSO for finding the near-optimal           many large-scale companies have begun considering
change plan.                                             environmental factors and evaluating their suppliers’
      The selection of green component suppliers in      environmental performance. He proposed a
this paper was completed in two phases. In phase one,    decision-making support tool to help companies
the component suppliers with a negative green index      integrate the environmental criteria into their supplier
were eliminated according to the WEEE and RoHS           selection process. Handfield [13] used the Delphi
directives. In phase two, an optimal decision-making     Method to interview 500 subjects from different
model was constructed and suppliers were selected by     companies, all experts on environmental management,
considering a quantity discount mechanism under the      concerning the significance of the environment
scenario of seeking the minimum total cost. The          performance index. He introduced environmental
solution of the model was obtained by applying PSO.      aspects into the procurement decision-making, and
Lastly, this study established a decision support        made the decision-making process more complex
system based on the solutions for the reference of       with quantitative and qualitative factors. Thus, this
decision-makers.                                         study proposed a decision-making support tool to
      The remainder of this paper is organized as        allow companies to integrate environmental
follows: Section 2 is the literature review on the       principles into their supplier selection process.
selection of green suppliers, WEEE and RoHS              Vachon & Klassen [30], using the printing industry in
directives, and particle swarm optimization. Section 3   Canada and the US as examples, have suggested the
proposes the assumptions and establishes the optimal     positive influence of partnerships in green plans on
mathematical model for green supplier selection.         costs, quality, delivery dates, flexibility, and
Section 4 details the flow of solution model             environmental issues. Noci [23] suggested that there
PSO. .Section 5 introduces the case study of product     are three steps in the selection process of green
design change of HDD and presents the                    suppliers: (1) determining the applicable green
decision-making system interface. Section 6 provides     strategies of the company (2) defining the operational
the conclusions and suggestions for future studies.      measuring method to assess the implementation of
                                                         environmental protective measures taken by the
                                                         suppliers and (3) selecting the most effective method
                                                         to proceed with the suppliers’ selection to ensure
22              International Journal of Electronic Business Management, Vol. 8, No. 1 (2010)

adherence to corporate environmental production         incorporates WEEE and RoHS in the EU directives
strategies. Che [2] provided optimal mathematical       for the selection of green partners when establishing
methods to help decision-makers select green            a supply chain. Thus, this study also included
partners in a balanced or defective supply chain        WEEE /RoHS in its green supplier selection.
network. Thus, this study also targeted green
suppliers in the planning of product change designs.    2.3 Particle Swarm Optimization (PSO)
                                                               Particle Swarm Optimization (PSO) was
2.2 WEEE and RoHS Directives                            proposed by Kennedy & Eberhart in 1995. Its
      Advancements in electronics and market            overall framework of evolutionary computation and
expansion have gradually replaced earlier electronic    basic concept was based on the simulation of
devices, which has led to increases in equipment        foraging of birds or fish. This optimization
waste and new environmental challenges. In 1998,        algorithm upon the group intelligence of living
Western Europe produced 600 million tons of             organisms was thus developed by observation of
electronic waste, and that quantity has increased by    group behaviors. In PSO, individuals are defined as
3-5% each year [9]. In 2004, over 300 million           “particles”, and optimization is achieved upon
computers in the US reached the end of their product    message exchanges among the group and the
life. In 2000, the WEEE Directive was divided into      experiences of the particles so that particles could
the current WEEE and RoHS Directive, which              move toward all possible areas in the space.
emphasizes recycling, reuse, and regeneration, but            PSO is a simple and usable method that is
also specifies the requirements on prohibited product   widely used to obtain optimal solutions in
materials.                                              optimization problems [34]. PSO is applicable for
      Willems [33] quantified the time of product       solving such complicated problems and may obtain
decomposition and fulfilled the economic feasibility    near-optimal solutions effectively. Wang and Wang
of systematic product decomposition. In this paper,     [32] indicated that PSO has advantages of cluster
the LP model was proposed to select the ultimate        intelligence and rapid iteration convergence, does not
strategy for optimal product life cycle and to study    involve function derivation when solving non-linear
the effect of decreasing expectation on disassembly     model, and has no ad hoc request on the number of
time and cost. A cost effective method and              equations and variables. As a result, PSO is able to be
optimization was reached through the view for final     widely used to solve non-linear models. Yu [35]
treatment of products. Although the LP model            pointed out that PSO is easy to use, is capable of
suggested that it was difficult to reach the            searching the global space, and is applicable for
optimization of ultimate product life decomposition     solving     complicated     non-linear   models of
in WEEE, the reduction of decomposition time on         optimization problems. The PSO algorithm, one of
medium and large products would have optimal            the optimization algorithms, has the following
results.                                                advantages [7]:
       According to new laws and regulations of         (1) The algorithm can be encoded with very simple
Europe, Queiruge [24] stated that recycling plants           programs, and can be applied parallel to
were built in Spain. A method to rank municipalities         numerous computations.
in Spain is proposed according to their adequacy of     (2) There are few parameters or predetermined
establishing recycling plants, and the Preference            transfer processes.
Ranking Organization Method for Enrichment              (3) It offers a rapid and accurate search capability
Evaluations was used for ranking. The plants were            when applied to optimization issues.
located in northern and eastern Spain; however, the
top ranking locations were southern and middle                Thus, PSO is a kind of algorithm based on
Spain. The method did not provide an optimal            active cooperation and message exchanges. This
mechanism for future recycling plants, but it did       simple concept is easily executed, and could be
offer optimal choices for selecting potential           converged immediately. PSO has been successfully
recycling plants. Hammond & Beullens [11]               applied to varied optimization issues [21,10,8,1,
acquired the balance under regulations from a           18,15]. Thus, this study applied PSO to obtain the
supply chain, in which supply was unable to meet        optimal solution.
demand in the closed chain. Their model involved              The calculation process of PSO is explained as
the WEEE directive, and the findings suggested          the following:
simulating reverse logistics of the supply chain,       Step 1: Randomly produce the velocity and position
based on the recycling rate of minimized new                     of each particle in n-dimension space.
products. Che [2] explored problems concerning          Step 2: Assess the fitness function value of each
production and delivery in a green supply chain, and             particle for the designed objective function.
constructed an optimal mathematical model to            Step 3: Compare the fitness function value with the
provide solutions. The mathematical model                        optimal function value of the particles in
Z. H. Che et al.: A Supplier Selection Model for Product Design Changes                                                                                23

        memory; the particles adjust the velocity for                                     i from supplier j for product p
        the next search, according to the optimal                                         Direct labor production cost of product
                                                           BFPP
        memories.                                                                         p
Step 4: Compare the individual optimal function                                           Quantity of component i from supplier j
        value with global optimal function value; if       D ip, j
                                                                                          for product p
        the individual value is higher than the global     p                              Discount quantity of component i from
        value, adjust the memory of global optimal             rqicnj
                                                                   ,
                                                                                          supplier j for product p at cn stage
        function value; each particle adjusts the          PDD                            Total demanded quantity of product
        velocity for the next searching, according to
        its optimal memory.                              Optimal mathematical model:
Step 5: Change the velocity and position of              Objective function:
        particles.                                           Fixed cost is added with variable costs to seek
Step 6: Stop when meeting the end condition,             the minimum objective function.
        otherwise, repeat Step 2; end condition is
        usually the optimal global value or                Minimize
        maximum generation value.                                               P                                       P
                                                                                                                                                     (1)
                                                                   z        TFC
                                                                             p 1
                                                                                                 p    PS      p       TVC
                                                                                                                       p 1
                                                                                                                                      p    PS   p


         3. OPTIMAL MODEL OF
         COMPONENT SUPPLIER                                  The variable cost refers to the total costs of
                                                         direct material costs, direct labor production costs,
        SELECTION OF PRODUCT                             and other manufacturing costs.
             DESIGN CHANGE
                                                                                     I       J
                                                          TVC          p         FC               i, j
                                                                                                     p       D ip, j  u ip, j
    The assumptions are described as follows:                                       i 1 j 1                                                        (2)
(1) The demand quantity for finished product is                                           BFP p  PDD  MCK                      p

     known and constant.                                  for all p
(2) Suppliers only accept orders with quantity less
     than the maximum productivity.                      Constraints:
(3) Transportation cost and losses are not               Select only one product.
     considered.
                                                               P


       The symbols in the mathematical model in this
                                                           
                                                           P 1
                                                                       PS       p         1                                                         (3)
research are as follows:
   p         Product Number, p  1,2,3,...P                  The suppliers’ quantity discount for different
  P          Total number of products                    quantities:
  i          Component Number, i  1,2,3,..., I
  I          Total number of components                                        p C i0, j ,                  0  D ip, j  p rq i1, j
                                                                              p 1
   j         Supplier Number, j  1,2,3,...J                                   C i, j ,                      p
                                                                                                                  rq i1, j  D ip, j  p rq i2, j
  J          Total number of suppliers                                        
                                                          FC       i, j
                                                                   p         .                                                                     (4)
             Hierarchy of quantity discount,                                  .
  cn
              cn  0,1,2,3,..., CN                                            
  CN         Total hierarchy of quantity discount                              p C icnj ,                    p
                                                                                                                  rq icnb  D ip, j  p rq icnj 1
                                                                                     ,                                ,                     ,

              1     Selection of product plan p            for all p,i,j
  PS p        
              0     Otherwise
                     Supplier j is selected for               The total supply quantity must be larger or equal
              1
  u ip, j           component i of product p            to product demand.
              0     Otherwise                                     P        I            J
  MCK p      Other manufacturing expenses of
             product p
                                                                  PS
                                                               p 1 i 1                 j 1
                                                                                                     p   D ip, j  PDD                               (5)

  TVC P      Variable cost of product p                            The supply quantity must be an integer.
  TFC p      Fixed cost of product p
              Direct    material   cost (purchase/
                                                               D ip, j  0, and  integer for all p,i,j
                                                                                                                                                     (6)
  p           transportation cost) of each unit of             PS p  0 ,1 for all p
      Cicnj
        ,
              component i from supplier j for product
              p upon quantity discount of cn stage
 FC ip, j     Purchase cost (each unit) of component
24                     International Journal of Electronic Business Management, Vol. 8, No. 1 (2010)

The suppliers’ supply quantity is between 0 and 5000.                                                         
                                                                           vih1  wvih  c1  rand   sipbest  sih         
                                                                                                                                      (9)
                                                                                       c2  rand   s     gbest
                                                                                                                      s
                                                                                                                       h
                                                                                                                       i   
     0D   i, j
                   5000 for all p,i,j,               (7)
           p                                                               sih 1  sih  vih 1                                     (10)

   4. SOLUTION MODEL OF                                               vih is the speed of particle i in generation h,
 GREEN SUPPLIER SELECTION
                                                               vih 1 is the speed of particle i in h+1 generation and
       OF COMPONENTS
                                                               w is the inertia weight, c1 and c 2 are Acceleration
      Each step of the proposed methodology is                 constants, s
                                                                              pbest
                                                                                    is the optimal position memory
shown in Figure 1 and is described step-by-step as                                       gbest
follows:                                                       value of the particle, s         is the optimal position
                                                                                                  h 1
Phase 1                                                        memory value of the group, si           is the position of
Step 1: Integrate WEEE and RoHS Directives
                                                               particle i in h+1 generation and rand () is the random
      WEEE and RoHS are integrated, as shown in
                                                               number between 0~1.
Table 1. The T-score of the suppliers corresponding to
                                                                Sub-step 3.4
WEEE and RoHS is transferred to the same unit, and
                                                                      After the second generation, compare the
summed. T formula is written as:
                                                               individual and group optimal values, and compare the
                                                               group optimal values to those of the previous
                    xx                                        generation.
       x                  50
                   x sd 10                                      Sub-step 3.5
                                                                      Repeat Sub-steps 3.2~3.4 until the iteration
             x' :    T score                          (8)      times are met.
             x:      primitive value                           Step 4: Obtain the acceptable product design
             x:      mean of x                                            change plan.
             xsd :   standard deviation of x

Step 2: Initial selection of suppliers                                  Integrate WEEE and RoHS Directives
      Arrange the total score of suppliers in order,
and eliminate the suppliers with inferior green index                                                                      Phase 1
according to the 80/20 principle.
                                                                            Initial selection of suppliers
Phase 2
Step 3: Use PSO to find the solution of the optimal
           model                                                         Use PSO to find the solution of the
      Find the fittest supply chain partnership                                   optimal model
combination and production delivery quantity based
on fixed cost, variable cost, and quantity discount.                         Generate random several
 Sub-step 3.1                                                                  sets of particles
      Randomly generate several sets of particles,                           Calculate fitnesses of all
according to existing market demand. Particles must                                                                        Phase 2
                                                                                    particles
meet all constraints; otherwise, randomly generate
another set until the pre-set number of particles is                          Compare fitness values
met.
 Sub-step 3.2                                                                Update the position and
      Calculate the fitness functions of all particles                         velocity of particle
for further comparison; to find the minimum
objective function, the minimum fitness function of                               Stop condition
all particles are the optimal values; store the optimal
values of the group for comparison with the later
generations.
                                                                        Obtain the acceptable product design
 Sub-step 3.3                                                                      change plan
      Update the positions and speed by (Eqs.
(9)~(10)), as proposed by Shi and Eberhart [27]                         Figure 1: Process of solution model
substituted the particles into the restraints for
calculation. If the result does not meet the restraints,
generate randomly until all updated results meet the
constraints. The later generation can thus be
calculated.
Z. H. Che et al.: A Supplier Selection Model for Product Design Changes                                       25

                                 Table 1: Integrated WEEE and RoHS directives
                      First hierarchy                              Second hierarchy
            Green image                        Customers' purchase or not (G1)
                                               Green customers' market share (G2)
            Green design                       Having recycling product design of supplier (G4)
                                               Having renewable product design of supplier (G5)
            Product recycling                  Product recycling rate (G6)
                                               Having reverse logistics system of supplier (G7)
            Green supply chain management      Passing ISO 14000 verification of supplier (G8)
                                               Having environmental protection policies of supplier (G9)
                                               Having environmental protection plans of supplier (G10)
            Pollution treatment cost           Solid wastes treatment costs (G11)
                                               Chemical wastes treatment costs (G12)
                                               Air pollution treatment costs (G13)
                                               Energy consumption costs (G14)
                                               Water pollution treatment costs (G15)
            Environment performance assessment Solid wastes treatment costs (G16)
                                               Energy consumption (G17)
                                               Air pollution (G18)
                                               Waste water (G19)
                                               Led content (G20)
                                               Mercury content (G21)
                                               Hexavalent chromium content (G22)
                                               PBB content (G23)
                                               PBDE content (G24)
                                               Cadmium content (G25)


   5. CASE APPLICATION AND                                   2500 units, including three change plans. Naturally,
                                                             other products, such as notebook, mobile phone, and
       SYSTEM INTERFACE                                      printer, could also be discussed by the proposed
                                                             approach. The concept of the product design change
     This study discussed a case of a single product,        for this illustrative example is shown in Figure 2.
with multiple change plans, multiple components,             Five components were required for each plan: upper
and multiple suppliers. It was assumed that each             casing (a), down casing (b), USB (c), chip (d), and
supplier has sufficient output to satisfy the demands,       leather pocket (e) (Figures 3-5); and five suppliers
thus one supplier is selected for each component.            could be selected for each component.
This case used External Hard Disk Drive (HDD)
enclosure as an example, with an order demand of

         product         plan    part :             part supplier combination

                                   A       1        2          3           4     5



                                   B       1        2          3           4     5



                          1        C       1        2          3           4     5



                                   D       1        2          3           4     5



                                   E       1        2          3           4     5
         External                                                                                 Solution
        Hard Disk                                                                                of product
          Drive           2                              ‧                                         design
         (HDD)                                           ‧                                       change of
                                                         ‧                                          HDD
                          3
                    Figure 2: The concept of product design change for External HDD enclosure
26                 International Journal of Electronic Business Management, Vol. 8, No. 1 (2010)

                                       down casing (b)                                                 down casing (b)
     upper casing (a)                                                upper casing (a)




      USB (c)                                 chip (d)                USB (c)                                  chip (d)
                        External HDD                                                    External HDD
                          enclosure                                                       enclosure


                      leather pocket (e)                                            leather pocket (e)


      Figure 3: Plan 1 for product design change                        Figure 5: Plan 3 for product design change

                                       down casing (b)                This study selected the change plans and
     upper casing (a)
                                                                component suppliers under the quantity discount
                                                                mechanism. It first selected the suppliers by WEEE
                                                                and RoHS upon the 80/20 principle. The original data
                                                                of WEEE and RoHS Directives of each supplier are
       USB (c)                                chip (d)
                                                                shown in Table 2. These data are transferred by the
                        External HDD                            formula (Occurrence value per month/Throughput per
                          enclosure                             month) collected from the companies’ business
                                                                databases, which recorded the historical data of each
                      leather pocket (e)
                                                                index for product manufacturing. Since the case had
                                                                the least number of suppliers, the 20% with inferior
                                                                performance were eliminated (Table 3). For example,
      Figure 4: Plan 2 for product design change                the supplier 1 is eliminated in Plan1, the supplier 2 is
                                                                eliminated in Plan 2, and the supplier 3 is eliminated
                                                                in Plan3 for part d. Then, the remaining entered phase
                                                                2.

                 Table 2: Partial list of original data of WEEE and RoHS Directives of each supplier
               Plan     Part     Supplier     G1         G2    G3     G4      G5         G6    G7        G8
                1        a          1          1           5     1      1       1         2      1         2
                1        a          2          1           4     1      1       1         2      1         2
                1        a          3          0           5     2      1       2         2      3         2
                1        a          4          1           3     1      1       1         2      1         1
                1        a          5          1           5     1      1       1         2      2         1
               Plan     Part     Supplier     G9         G10   G11    G12     G13       G14    G15       G16
                1        a          1          2          40   200    300     310       1012   313       312
                1        a          2          5          41   220    320     310       989    319       323
                1        a          3          3          41   210    315     321       990    322       325
                1        a          4          2          40   205    310     324       996    323       320
                1        a          5          2          39   200    305     315       992    325       316
               Plan     Part     Supplier     G17        G18   G19    G20     G21       G22    G23       G24
                1        a          1         501        314    90     18      15        12     18        12
                1        a          2         500        319    97     13      17        17     18        12
                1        a          3         516        316    98     12      16        12     14        14
                1        a          4         504        322    95     13      12        15     12        15
                1        a          5         519        310    98     17      17        18     18        18


         6. CONCLUSIONS AND                                     selected the suppliers by WEEE and RoHS, and used
                                                                the optimal model to select the suppliers with
             SUGGESTIONS                                        minimum total costs. The optimal model was
                                                                calculated by PSO. This study established a
      In product design change, it is critical that             decision-making support system based on this model
decision makers select component suppliers with                 and applied it to a case study. The results showed that
environmental maintenance abilities. This study                 with this system, decision makers could immediately
constructed a decision-making model for selecting               acquire the desired results by inputting related data
design changes and green suppliers. This model first            and parameters.
Z. H. Che et al.: A Supplier Selection Model for Product Design Changes                                                                                  27

               Table 3: The result of phase 1                             Some possible future research directions on
                   Plan 1       Plan 2          Plan 3              product design change could include: 1) accounting
      Part        Supplier     Supplier        Supplier             for transportation cost and loss in optimal model
       a             1            1               1                 development for plan assessment, 2) considering
                     2*           2*              2*
                     3            3               3
                                                                    optimal product-mix, level of stocking, and
                     4            4               4                 scheduling in mathematical development for
                     5            5               5                 multi-product production, 3) developing the
       b             1            1               1*                integrated mathematical model while considering the
                     2            2               2                 mechanism of critical information sharing between
                     3            3               3                 the manufacturer and supplier, 4) considering the
                     4            4               4
                                                                    uncertainty of the demand patterns, costs, and
                     5*           5*              5
       c             1            1*              1*                capacities, 5) applying other heuristic approaches for
                     2            2               2                 solving the problem and comparing with the PSO,
                     3            3               3                 and 6) enhancing the problem solving quality by
                     4            4               4                 modifying the proposed methodology for supply
                     5*           5               5                 chain design.
          d          1*           1               1
                     2            2*              2
                                                                                       520000
                     3            3               3*
                                                                                                                                      Plan 1
                     4            4               4                                    510000                                         Plan 2
                     5            5               5                                                                                   Plan 3
       e             1*           1               1*                                   500000
                     2            2               2                       Total Cost
                                                                                       490000
                     3            3*              3
                     4            4               4                                    480000
                     5            5               5
                                                                                       470000
    *: the supplier with inferior performance is eliminated.
                                                                                       460000
                                                                                                1   9   17   25 33   41 49 57      65 73   81 89    97
                                                                                                                     Generations

                                                                        Figure 6: The convergence processes of three plans.

                                                               Result
          Input

   Customer
   demand                                                                                                                           Selected
                                                                                                                                    plan
   Particles
   number                                                                                                                          Supplier No.
                                                                                                                                   for part a
   Generation
   number                                                                                                                          Supplier No.
                                                                                                                                   for part b
   Speed
   constraints                                                                                                                     Supplier No.
                                                                                                                                   for part c
    Weight
                                                                                                                                   Supplier No.
    Run                                                                                                                            for part d
                                                                                                                                   Supplier No.
                                                                                                                                   for part e
                                                                                                                                   Total cost for
                                                                                                                                   the selected
                                                                                                                                   plan


                                 Figure 7: Interface of decision-making support system

           ACNOWLEDGMENTS                                                                           REFERENCES
       This research was partially financially                      1.                 Amitava, C., Ranajit, C., Fumitoshi, M. and
supported by the National Science Council under                                        Takahiro, E., 2007, “Neuro-fuzzy state
project No. NSC 96-2416-H-027-MY2 and No. NSC                                          modeling of flexible robotic arm employing
95-2221-E- 027-078. We also wish to thank the                                          dynamically varying cognitive and social
Editor-in-Chief and the two anonymous referees for                                     component based PSO,” Measurement, Vol. 40,
their valuable comments.                                                               pp. 628-643.
28               International Journal of Electronic Business Management, Vol. 8, No. 1 (2010)

2.    Che, Z. H., 2009, “Using fuzzy analytic                   Management Network, pp. 28-30.
      hierarchy process and particle swarm                15.   Hu, X., Eberhart, R. C. and Shi, Y., 2003,
      optimization for balanced and defective supply            “Engineering optimization with particle
      chain problems with considering WEEE/RoHS                 swarm,” Proceedings of the IEEE Swarm
      directives,”      International    Journal     of         Intelligence Symposium 2003, Indianapolis,
      Production Research, DOI: 10.1080/00207540                Indiana, USA, pp. 53-57.
      802702080.                                          16.   Humphreys, P. K., Wong, Y. K. and Chan, F. T.
3.    Cooper, R. G., 1993, Winning at New Products,             S., 2003, “Integrating environmental criteria
      Addison Wesley Publishing CO., Boston.                    into the supplier selection process,” Journal of
4.    Cooper, R. G., Edgett, S. J. and Kleinschmidt, E.         Materials Processing Technology, Vol. 138, No.
      J., 1998, Portfolio Management for New                    1-3, pp. 349-356
      Products, Perseus Bools, New York.                  17.   Jonghoon, C. and Lee, K., 2002, “A framework
5.    Deans, I., 1999, “An approach to the                      of collaborative design environment for
      environmental management of purchasing in                 injection molding,” Computers in Industry, Vol.
      the utilities sector,” Eco-Manage Audit, Vol. 6,          47, No. 3, pp. 319-337.
      pp. 11-17.                                          18.   Kathiravan, R. and Ganguli, R., 2007,
6.    Dowlatshahi, S., 2000, “Designer-buyer-                   “Strength design of composite beam using
      supplier interface: Theory versus practice,”              gradient and particle swarm optimization,”
      International Journal of Production Economics,            Composite Structures, Vol. 81, No. 4, pp.
      Vol. 63, No. 2, pp. 111-130.                              471-479.
7.    Eberhart, R. C. and Shi, Y., 2001, “Comparing       19.   Krause, D. R., Mark, P. and Sime, C., 2001,
      inertia weights and constriction factors in               “Toward a measure of competitive priorities for
      particle swarm optimization,” Proceeding of               purchasing,”      Journal     of     Operations
      the     2000      Congress     of   Evolutionary          Management, Vol. 19, No. 4, pp. 497-512.
      Computation, Vol. 1, pp. 84-88.                     20.   Lai, E. F., 2003, “Green supplier management
8.    Elbeltagi, E., Hegazy, T. and Grierson, D., 2005,         and evaluation system: Taking LCD industry as
      “Comparison among five evolutionary-based                 an example,” Department of Resources
      optimization         algorithms,”      Advanced           Engineering, National Cheng Kung University,
      Engineering Informatics, Vol. 19, No. 1, pp.              Masters Thesis.
      43-53.                                              21.   Lin, Y. L., Chang, W. D. and Hsieh, J. G.., 2008,
9.    European Commission, http://www.eia.org/                  “A particle swarm optimization approach to
      download/eic/21/www Final Proposal June                   nonlinear rational filter modeling,” Expert
      2000.htm.                                                 Systems with Applications, Vol. 34, No. 2, pp.
10.   Fawaz, S. and Ali, A., 2007, “A self-adaptive             1194-1199.
      differential evolution heuristic for two-stage      22.   Noci, G., 1995, “Supporting decision making
      assembly scheduling problem to minimize                   on recycling based investments,” Business
      maximum lateness with setup times,” European              Strategy and the Environment, Vol. 4, No. 2, pp.
      Journal of Operational Research, Vol. 182, No.            62-71.
      1, pp. 80-94.                                       23.   Noci, G., 1997, “Designing ‘green’ vendor
11.   Hammond, D. and Beullens, P., 2007,                       rating systems for the assessment of a
      “Closed-loop supply chain network equilibrium             supplier’s     environmental      performance,”
      under legislation,” European Journal of                   European Journal of Purchasing & Supply
      Operational Research, Vol. 183, No. 2, pp.                Management, Vol. 3, No. 2, pp. 103-114.
      895-908.                                            24.   Queiruga, D., Walther, G., Gonzalez-Benito, J.
12.   Handfield, R. B., Ragatz, G. L., Petersen, K. J.          and Spengler, T., 2008, “Evaluation of sites for
      and Monczka, R. M., 1999, “Involving supplier             the location of WEEE recycling plans in
      in new product development,” California                   Spain,” Waste Manage, Vol. 28, No. 1, pp.
      Management Review, Vol. 42, No. 1, pp. 59-82.             181-190.
13.   Handfield, R., Steven, V. W., Sroufe, R. and        25.   Sarkis, J., 2003, “A strategic decision
      Melnyk, S. A., 2002, “Applying environmental              framework       for    green    supply     chain
      criteria to supplier assessment: A study in the           management,” Journal of Clean Production,
      application of the analytical hierarchy process,”         Vol. 11, No. 4, pp. 397-409.
      European Journal of Operational Research,           26.   Sirias, D. and Mehra, S., 2005, “Quantity
      Vol. 141, No. 1, pp. 70-87.                               discount versus lead time-dependent discount
14.   Hass, J. L., 1996, “Greening the supply value             in an inter-organizational supply chain,”
      chain: A case study and the development of a              International Journal of Production Research,
      conceptual model,” The Third Conference of                Vol. 16, No. 16, pp. 3481-3496.
      the      Nordic       Business     Environmental    27.   Shi, Y. and Eberhart, R. C., 1998, “A modified
Z. H. Che et al.: A Supplier Selection Model for Product Design Changes                                     29

      particle swarm optimizer,” IEEE International               ABOUT THE AUTHORS
      Conference on Evolutionary Computation, pp.
      4-9.                                                 Zhen-Hua Che received his Ph.D. in Industrial
28.   Shin, H. and Benton, W. C., 2007, “Quantity          Engineering and Management at National Chiao
      discount      approach     to     supply    chain    Tung University. He is an Associate Professor at the
      coordination,”      European        Journal     of   Department of Industrial Engineering and
      Operational Research, Vol. 180, No. 2, pp.           Management, National Taipei University of
      601-616.                                             Technology, R.O.C. His research interests include
29.   Tsai, J. F., 2007, “An optimization approach for     production/operations management, supply chain
      supply chain management models with quantity         management, and information management.
      discount policy,” European Journal of
      Operational Research, Vol. 177, No. 2, pp.           Tzu-An Chiang is an assistant professor in the
      982-994.                                             Department      of    Commerce      Automation    &
30.   Vachon, S. and Klassen, R. D., 2006, “Green          Management at National Pingtung Institute of
      project partnership in the supply chain: The         Commerce (NPIC), Taiwan. He received his doctoral
      case of the package printing industry,” Journal      degree in the Industrial Engineering and Engineering
      of Clean Production, Vol. 14, No. 6/7, pp.           Management from National Tsing Hua University. Dr.
      661-671.                                             Chiang’s research interests are in the areas of R&D
31.   Wang, H. S. and Che, Z. H., 2009, ”Applying          management, data mining, and production
      and comparing four different PSO approaches          management.
      in integrated problem of production change
      planning, part supplier selection, and quantity      Chuang Tu received his Ph.D. in Business and
      allocation,” Journal of the Chinese Institute of     Management at National Chiao Tung University. He
      Industrial Engineers, Vol. 26, No. 2, pp. 87-98.     is an Associate Professor at the Department of
32.   Wang, S. T. and Wang, Z. J., 2005, “Study of         International Business and Administration, Chienkuo
      the application of PSO algorithms for nonlinear      Technology University, R.O.C. His research interests
      problems,” Journal of Huazhong University of         include service management, quality management,
      Science and Technology, Vol. 33, No. 12, pp.         and performance management.
      4-7.
33.   Willems, B., Dewulf, W. and Duflou, J. R.,           Cheng-Jui Chiang received his Master in Industrial
      2006, “Can large-scale disassembly be                Engineering and Management at National Taipei
      profitable? A linear programming approach to         University of Technology. He is a Ph.D. student at
      quantifying the turning point to make                the Department of Industrial Engineering and
      disassembly          economically         viable,”   Management, National Taipei University of
      International Journal of Production Research,        Technology, R.O.C. His research interests include
      Vol. 44, No. 6, pp. 1125-1146.                       production/operations management, supply chain
34.   Yu, F. H., Liu, H. B. and Dai, J. B., 2006,          management, multi-objective optimization, and
      “Grey particle         swarm algorithm for           algorithms.
      multi-objective       optimization     problems,”
      Journal of Computer Applications, Vol. 26, pp.       (Received January 2009; revised May 2009; accepted
      2950-2952.                                           June 2009)
35.   Yu, H. J., Zhang, L. P., Chen, D. Z., Song, X. F.
      and Hu, S. X., 2005, “Estimation of model
      parameters using composite particle swarm
      optimization,”       Journal     of     Chemical
      Engineering of Chinese Universities, Vol. 19,
      No. 5, pp. 675-680.
30      International Journal of Electronic Business Management, Vol. 8, No. 1 (2010)


                  產品設計變更之供應商評選模式
                   車振華 1、江梓安 2、杜壯 3、蔣承叡 1*
                   1
                     國立台北科技大學工業工程與管理系
                       台北市忠孝東路三段一號
                2
                  國立屏東商業技術學院商業自動化與管理系
                        屏東市民生東路 51 號
                          3
                            建國科技大學
                        彰化市介壽北路一號

                                         摘要
     環境污染問題已經嚴重危害地球生態環境並成為全球共同議題,綠色產品已成為企業
     滿足市場多元化需求與提升產品競爭力之首要趨勢。鑑此,產品變更設計時之符合環
     保要求之綠色零組件供應商評選成為企業生產新產品之主要先期作業。本研究以綠色
     零組件供應商評選為標的,先以歐盟 WEEE 與 RoHS 指令為依據剔除於綠色指標表現
     較差之零件供應商,並再建構一最佳化決策模式,於考量數量折扣機制及追求總成本
     最小化情境下進行供應商評選。最後,本研究利用粒子群演算法( Particle Swarm
     Optimization,PSO)進行模式求解,並將決策過程建構成一決策支援系統,以供決策
     人員使用。

     關鍵詞:產品設計變更、供應商評選、數量折扣、粒子群演算法
     (*聯絡人:wind112233@yahoo.com.tw)

								
To top