- Email: [email protected]
INTEGRATED FRAMEWORK FOR REVERSE LOGISTICS SYSTEM: IMPROVED DECISION MAKING PROSPECTIVE
INTEGRATED FRAMEWORK FOR REVERSE LOGISTICS SYSTEM: IMPROVED DECISION MAKING PROSPECTIVE
Jitendra Madaan, Subhash Wadhwa & Mudit Verma* Indian Institute of Technology, New Delhi, India *MSIT, New Delhi, India Email: [email protected] Abstract: Direct sales channels, environmental legislation, and liberal returns policies have contributed to the growth of returns flows. However there is little empirical research related to the modeling of an integrated framework for returns. Till today, most of forward supply chain models focus only on physical transactions until the product reaches the customer’s end, leaving key decision variables implicit for some or all of the parties involved for return of product back from customer to the manufacturer. In this paper, we propose an alternative approach that explicitly addresses the extension of supply chain model for reverse logistics. As a matter of fact framework for decision modeling reverse logistics barely exists to support return. Therefore in order to fill this gap we discuss over the key decisions related with return handling and identifying quantitative models to support those decisions. Copyright © 2007 IFAC Key words: reverse logistics, decision making, and product recovery.
1. INTRODUCTION Conventional forward supply chain modeling usually considers a set of processes, driven by customer demand, that convey goods from suppliers through manufacturers and distributors to the end customers. But, this is not where the value of physical product terminates. Physical goods as well as their values do not simply get consumed fully once they have reached the customer. In order to capture this value we require a broadening of the supply chain perspective to include new processes, known as ‘reverse logistics’, and multiple interrelated usage cycles that are linked by specific market interfaces. Coordinating the successive activities after the product return is a key to maximizing the value generated. Therefore, many goods/products move beyond the conventional supply chain horizon, thus triggering additional business transactions: used products are sold on secondary markets; outdated products are upgraded to meet latest standards again; failed components are repaired to serve as spare parts; unsold stock is salvaged; reusable packaging is returned and refilled; used products are recycled into raw materials again. The set of processes that accommodate the flow of these goods ‘upstream’ in a conventional supply chain scheme, are known as ‘reverse logistics’. Today reverse logistics for returns have become endemic even in technologically advanced products, with rates as high as 20% in some sectors. Therefore, developing a comprehensive and cost-effective approach to handling returns is a daunting challenge that reaches well beyond the operational level. Thus, a well-developed reverse logistics and management plan can be a vital strategic asset. Further, supply chain management analyses have already revealed that this type of product returns can, in fact, be beneficial for both the manufacturer and the retailer. Thereby, the manufacturers benefit hinges on larger expected sales volume, competition marketing motives, direct economic motives, and concerns with the environment. Some of these are shown in figure 1.
655
Figure 1 Benefits Extracted From Reverse Logistics With the legislative measures tightening up, there are not many options left with the companies but to go for reverse logistics practices. According to a survey conducted by the Reverse Logistics Executive Council (RLEC), the average returns rate is 8.46% with individual expected return shown in table 1. Table 1: Expected Rate of Return (Survey by RLEC) Product category Return % in Year 2004 White goods 8% House Hold appliances 7% TV’s 8% Computers and 15% accessories Brown Goods 6% But looking across the entire manufacturing value chain, one finds return rates are as high as 20% or more in the year 2004 and these rates are expected to increase more in near the future. Therefore there is a need to achieve efficiency in resource consumption, reduction in waste and extraction of value from the return, when there in one out of five products is being returned. Thus in a closed loop system it will be necessary to develop new approaches for the systematic use and re-use of
resources in a broader sense. This paper presents a new approach for extending supply chain modeling framework for reverse logistics in order to capture opportunities for improved decision making. 2. MODELING FRAMEWORK FOR INTEGRATED REVERSE LOGISTICS SYSTEM Till today, most often companies have not even mapped processes related to them (Stock, 1998). Returned products are still treated on an ad hoc basis and regarded as waste. The lack of interest can partly be explained through the fact that managing product returns flows is complicated due to their interfunctional and inter-enterprise nature. Several methodologies like business process reengineering (BPR), enterprise modeling (Sgegheo & Andersen, 1999; Whitman & Huff, 2001), enterprise integration modeling (Petrie, 1992; Presley, 1997), and integrated enterprise modeling (IEM) have been suggested to improve enterprise performance through the designing and modeling of business processes. All these efforts concern themselves with the configuration of processes to achieve enterprise goals, with slightly different focuses in forward direction, but a lot hasn’t been done for modeling the framework with product return. RES (Reverse Enterprise System) requires enterprise and inter-enterprise relationships are viewed or modeled from return perspectives required which are to be fully described and understood in forward direction. Traditional supply chain modeling methodologies typically emphasize forward aspects keeping in view the chain ends as products reaches to customers end (Rogers & Tibben, 1999 & Thierry, 1995). However, here in this paper we look forward to model product return aspect from vastly different conceptual perspectives. Return process can be looked at from various perspectives depending on the kind of information required. Usually this will represent the activities comprising: when product came back into the chain; what kind of re-manufacturing functions are required to be done; who is performing these activities; how and why are they executed; when and where are these activities performed; and what data elements is required to be manipulated. Looking at the prior research, reverse logistics has multiple views or perspectives such as function (evaluation or screening, remanufacturing, repair, resell etc), information (related to product usage and return), behavior (uncertainty), organization (structure of organizing the return channel), resource (control the flow product in return chain)
656
Figure 2 Business Model for Strategic Reverse Enterprise System ,business and legislative rules, decisions, goals, and actors (or people) views for the return process. However, the existing methodologies that have been utilized for the modeling and integration of an enterprise do not specifically address the techniques for inter and intra reverse enterprise system modeling and integration. Therefore this paper, presents a generic model to support understanding, analysis, and design of product return processes aligned with information systems for inter-enterprise integration between enterprises as well as for a particular enterprise. We start with development of generic product recovery architecture for a reverse enterprise system as shown in figure 3.
Figure 3 Generic Product Recovery Architecture for a Reverse Enterprise System The development of product recovery architecture includes an integrated modeling framework and a suite of modeling techniques for enterprise modeling. The framework and methodology provide a systematic means to simultaneously achieve three kinds of integration: model integration of the different views, paradigm integration of the traditional and object oriented paradigms, and the integration of the reverse enterprise systems. The methodology here includes acceptance criteria which determine whether a material is suitable for specific uses or activities. These criteria are applied within the model to guide selection of materials for specific applications and their successive uses. 3. OPPORTUNITIES FOR IMPROVED DECISION MAKING IN REVERSE LOGISTICS Decision making process can be defined as a methodical approach of decision making that allows managers to handle problems where different alternatives and/or a certain degree of uncertainty are involved. Product return is well known for its uncertainty; therefore main objective of decision analysis in reverse logistics is to propose a model to assist decision makers in making better decisions. The decision problem for reverse logistics can be effectively solved just by using an appropriate decision making method. But, there are number of decision-making methods mentioned in the literatures [Stock, 1998]. The availability of a wide variety of decision-making
methods leads to another problem of choosing a suitable method. These decision-making methods differ widely in the purposes they serve, their ease of use and theoretical soundness, and the evaluations they yield. An intending user must thus consider the appropriateness of the method to the problem in terms of the value judgments it asks from the decision maker, the types of alternatives it can consider, and the forms of evaluations it yields. Although there have been some models developed for the evaluation of product after return options, they are either mono criterion- or bi-criteria based. Some of these efforts are focused only on some of the aspects or on one or two product return processes. The economics of design for re-cycling by using cost and benefit analysis method was given by Kroon & Vrijens [1995]. By this method the cost of each return option was first computed, followed by the calculation of the benefits of each of the options. The results of the cost and benefits calculations of each of the options were compared to determine the most profitable alternative. However, the focus of the work is on the product design with the consideration of the value alone, excluding the utilization stage. Furthermore, the return options considered are parts reuse/resale, product recycling, cannibalization, and remanufacture/ repair etc. while the basis of evaluation is limited to environmental and economic factors. Low and associates presented a number of mathematical models to assist designers in evaluating a number of return (reverse manufacturing) options of a product at the conception stage of the product development. The options being considered are recycling, remanufacturing, resale, repair, and disposal etc. The cost models evaluate the cost of each model as a fraction of the reverse manufacturing cost and consequently evaluate the trade-off between the options. Again the basis of evaluation is only financial and is directed at the product design. Moreover, authors have also reported a number of efforts on remanufacturing and disassembly [Fleischman, 2000, Giultinian, 1975, Krikke et al., 1999, Kroon & Vrijens, 1995 and Dowlatshahi, 2000]. Among them is the development of metrics for the assessment of remanufacturability of designs and for measuring ease of assembly, disassembly, testing, inspection, cleaning, and part replacement by Bras and associates. Here in this paper our model includes the both cost and utility based aspects of the product return process being studied. 4. DECISION MAKING APPROACH FOR REVERSE LOGISTICS This paper focuses on manager’s decision for reverse logistics process for product return. Problematic areas or challenges in the current product return practices were identified. These challenges were described in section 1 and 2. Based on the challenges described, the decision making process was studied to minimize the problems that could be encountered. The quality and quantity logistics decisions define the choice of reverse manufacturing process. Barry et al. [1993] and Carter& Ellram [1998] identified a number of process alternatives for product recovery. The five 657
notable ones among these reverse manufacturing techniques are repair and maintenance, refurbishing; remanufacturing; cannibalization, and reuse. The decision about each of these alternatives is initiated by functional level breakdown and flow block analysis. It is followed by the assessment of resource requirements for reverse logistics operation, and the data resulting from these are used to evaluate the suitability of the reverse manufacturing option. Figure 3 shows various reverse manufacturing options from firm to firm both for a particular product and for different products, a particular scenario can be chosen as per the availability of resources. 4.1 Evaluation and analysis of Decision for reverse manufacturing Various decision evaluation approaches and applications such as [Demmel, J.G. and Askin, R.G. (1986), Sanchez et. al (1994)] revealed that a product value function can be used to describe a relationship between a set of attributes of same dimension of value and the degree of utility corresponding to that attribute. Here the proposed value function can be applied to decision making process for reverse manufacturing function. The value or utility of each option can thus be calculated as a measure of preference for various values of a variable, having measured the relative strength of desirability that the decision maker has for those values. Suppose {α, α1, α2, α 3, α m} are the feasible reverse manufacturing function for the decision problem (represented by Figure 3), {β1, β2, βn} is a set of attributes, and δmn denotes a specific level of βn with regard to reverse manufacturing option αm. Here if we follow decision theory then certain preferential and independence conditions hold, then υ (δ11, δ12, δmn) has the form of a simple additive weighted utility value function: N
υ j (α ) = ∑ wiυ i (δ i ) i =l
Where υi(δi) = Value function for a attribute βi wi = weight-age given to the attribute βi υj(α) = The utility value of reverse manufacturing alternative αj on attributes ∑ {β1, β2, ..., βN} = The summation of the utility value at each of the attributes These generalised value functions can be rewritten for each reverse manufacturing alternative as follows: Repair process value function α 1 = w1[γ1(Dcost + O cost)]1 + w2 [γ2 (Sop + P Ch + Tcap)]1 + w3[γ3 (Rcon + Wrl + Wi)]1 + w4 [γ4 (Rc + Dr + Rr)]1 + w5[γ5 (Tr + Ts + To + T dly + Taux)]1 Refurbishing process value function α 2 = w1[γ1(Dcost + O cost)2 +w2 [γ2 (Sop + P Ch + Tcap)]2 + w3[γ3(Rcon + Wrl + Wi)]2 + w4[γ4 (Rc + Dr + Rr)]2 + w5[γ5 (Tr + Ts + To + T dly + Taux)]2 Remanufacturing process value function α 3 = w1[γ1(Dcost + O cost)3 +w2 [γ2 (Sop + P Ch + Tcap)]3 + w3[γ3(Rcon + Wrl + Wi)]3 + w4[γ4 (Rc + Dr + Rr)]3 + w5[γ5 (Tr + Ts + To + T dly + Taux)]3 Cannibalization process value function
α 4 = w1[γ1(Dcost + O cost)4 +w2 [γ2 (Sop + P Ch + Tcap)]4 + w3[γ3(Rcon + Wrl + Wi)]4 + w4[γ4 (Rc + Dr + Rr)]4 + w5[γ5 (Tr + Ts + To + T dly + Taux)]4 Reuse process value function α 5 = w1[γ1(Dcost + O cost)5 +w2 [γ2 (Sop + P Ch + Tcap)]5 + w3[γ3(Rcon + Wrl + Wi)]5 + w4[γ4 (Rc + Dr + Rr)]5 + w5[γ5 (Tr + Ts + To + T dly + Taux)]5
Where Dcost ---- Direct costs O cost ---- Overhead cost Sop ---- Product state after return P Ch ---- Process characteristics Tcap ---- Techno-capability Rcon ---- Resource consumption Wrl ---- Waste released Wi --- Waste impact Rc ---- Resources conserved Rr ---- Resources required to reprocess Dr ---- Demand of reprocessed product Tr ---- Time Reverse manufacturing option Ts ---- Set-up time To ---- Actual process operations time Taux ---- Auxiliary times T dly ---- Delay time γ 1 ---- Normalizing for cost attribute δ 1 γ 2 ---- Normalizing for technical attribute δ 2 γ 3 ---- Normalizing for environmental attribute δ 3 γ 4 ---- Normalizing for market attribute δ 4 γ 5 ---- Normalizing for time attribute δ 5
percentage – change-based analysis is applied by changing the estimates in increments of plus and minus ten percent and recompiling the results. Consequences of the uncertainty in the accuracy of data for evaluation, sensitivity analysis of how variation in data affects the overall performance of a reverse enterprise system and the effect of decision making process for various reverse manufacturing & logistics functions shown in fig 4 & fig 5.
4.3 Improved Decisions After assessing each reverse manufacturing alternative on the all attributes, the results have to be compared with the satisfaction of minimum standard on each of the attributes. The final selection of the alternative to be used for the extension of a particular product in a specific location can be based on three principles, namely: satisfying solution, maximization of expected utility, and preferred solution [Sanchez et al., 1994]. 4.4 Minimum acceptable solution The set of satisfying solutions consist of all processes that meet minimal requirements: l ⎡ l ⎤ aia = {ai ∑ wi vi ( f i ) ≥ ⎢∑ wi vi ( f i )⎥ } i =1 ⎣ i =1 ⎦ min
⎡
⎤
l
∑ w v ( f )⎥⎦ min ⎣
Where ⎢
4.2 Sensitivity Analysis
i i
i =1
i
= Minimum acceptable
performance.
Sensitivity analysis, which refers to the study of how important results changes with changes in estimates, is a “what-if” technique that looks at how a result will be changed if assumptions change or original estimates are not achieved. It is applicable in any analytical technique involving uncertainty in their underlying assumptions.
4.5 Maximum benefit solution This decision is for a decision-maker in favor of maximizing expected utility/benefit. In this case, recourse is not made to minimum satisfactory condition level with respect to any attribute. Thus this solution is purely based on compensatory method that permits tradeoffs between the attributes. Nj
amb = {ai max ∑ w f vi f ( f i ) j =1
⎡ Nj ⎤ ⎢∑ w j ⎥} ⎣ j =1 ⎦
4.6 Preferred solution These are the solutions which are both acceptable and benefit maximizing. The solutions are the one utilizing the integration of both compensatory and noncompensatory techniques, combining the advantages of the methods. This means, selecting the reverse manufacturing option so that:
Figure 4 Sensitivity of reverse manufacturing option
with respect to cost can either be Numerical sensitivity analysis
Nj
It is recognized as an aid for validating the model and for identifying model improvement possibilities [Sanchez et al., 1994]. Sensitivity analysis may be carried out numerically or by differentiation. Numerical sensitivity analysis can either be displayed as absolute amounts or as percentage changes from the base estimates or both. In this paper, the 658
aideal = {ai max ∑ w f vi f ( f i ) j =1
⎡ Nj ⎡ Nj ⎤ ⎢ ∑ w j ⎥} ≥ ⎢ ∑ w f vi f ( f i ) ⎢⎣ j =1 ⎣ j =1 ⎦
⎡ Nj ⎤ ⎤ ⎢∑ w j ⎥ ⎥ ⎣ j =1 ⎦ min ⎦⎥
Depending on the nature of the decision maker, represented by the three mentioned decision making principles earlier, substituting all the relevant values obtained from equations 3.3 - 3.7 and equations 2.65 -
2.71 into any of equations 3.8 - 3.10 results in an optimal reverse manufacturing process selection. 5.0 COMPUTER BASED IMPLEMENTATION OF DECISION MODEL The comprehensiveness of this model and the data requirements with the attendant calculations and analyses make the application of the methodology tedious without the use of a computer. A Computer application does not only quicken the implementation of the model but also facilitates easy and fine presentation of the implementation results. This model can be easily implemented on a computer by using any of the windows application programmes such as Visual Basic, Visual C++ and others. However, ARENA 7.0 simulation package is used in this work to develop the demonstrative computer implementation prototype. The prototype can later be upgraded to a decision support tool for reverse enterprise systems. This demonstrative computer prototype also supports the decision model in assessing other parameters like the life-extendibility of the retired product, marketability of the reprocessed product and the cost of adopting a specific process in extending the life of the retired product. Furthermore, it facilitates the evaluation of available facility’s suitability for the process and consequently for the chosen reprocessed product quality. The process time, and the conformity of the process to legislative requirement can equally be determined by using the computer application prototype. 5. CONCLUSION Despite all the research efforts thus far being put into resource recovery and reuse, none has been found to consider their decision-making aspects in a comprehensive manner (Wadhwa & Madaan, 2004). Most of the research paper focuses on the development of some aspect of reverse manufacturing techniques in the areas like product take-back logistics. Others focus on product design that facilitates resource recovery and reuse. The absence of a comprehensive decision making framework in the area of resource creates a void which this paper is attempting to fill. This paper proposed an integrated framework and model for improved decision making for reverse manufacturing option selection methodology, and a proposed scope for implementation of computer based implementation of the methodology. This paper, by establishing parameters needed for the evaluation of reverse manufacturing and logistics processes; by developing a suitable correlations for decision making in this domain; as well as by developing framework for setting minimum standard on major decision making parameters and demonstrating its applications, will be useful for decision makers. REFERENCES Barry, J.; Girard, G. and Perras, C. (1993) Logistics planning shifts into reverse. Journal of European Business, 5(1), 34–38. Carter, C.R. & Ellram, L.M. (1998) Reverse Logistics: A review of the literature & framework for future investigation, J. of Business Logistics, 19(1), 85102. 659
Demmel, J.G. and Askin, R.G. (1986). A multiple objective decision model for the evaluation of advanced manufacturing system technologies. J. Manuf. Systems 11(3), 133-152 Dowlatshahi, S. (2000) Developing a theory of Reverse Logistics. Interfaces, 30(3), 143-155. Fleischman M.; Krikke H.R.; Dekker R. and Flapper S.D.P. (2000) A characterization of logistics networks for product recovery. Omega, The International Journal of Management Science, 28(6),pp 653- 666. Giultinian, J.P. and Nwokoye, N.G. (1975) Developing distribution channels and systems in the emerging recycling industries. International Journal of Physical Distribution, 6(1), 28-38. Hillegersberg, J.; Zuidwijk, R.; Nunen, J. and Eijk, D. (2001) Supporting Return Flows in the Supply chain. Communications on the ACM, June, 44(6),pp 74-79. Krikke, H.R.; Harten, A. and Schuur, P.C. (1999) Business case Roteb: Recovery strategies for monitors. Computers and Industrial Engineering, 36(4), 739-757. Kroon L., Vrijens G. (1995). Returnable containers: an example of Reverse Logistics. International Journal of Physical Distribution and Logistics Management, 25(2), pp 56-68 Petrie, C. (Ed.), (1992). Introduction. First International Conference on Enterprise Integration Modeling, Austin, TX. Pohlen, T.L. and Farris II, M.T. (1992) Reverse Logistics in plastics recycling, International Journal of Physical Distribution and Logistics Management, 22(7), pp 35-47. Presley, A. R (1997). A representation method to support enterprise engineering. PhD Dissertation, Industrial and Manufacturing Systems Engineering, University of Texas at Arlington Logistics Executive Council. Reverse http://www.unr.edu/coba/logis Ritchie, L.; Burnes, B.; Whittle, P. and Hey, R. (2000) The benefits of Reverse Logistics: the case of Manchester Royal Infirmary Pharmacy. Supply Chain Management: an international Journal, 5(5), pp 226-233. Rogers, D.S. and Tibben-Lembke, R.S. (1999) Going backwards: R L trends & practices. Reverse Logistics Executive Council, Pittsburgh, P.A. Sanchez, J.M.; Priest,J.W. and Burnell, L.J. (1994). Design Decision Analysis and Expert systems in Concurrent Engineering. In: Dorf, R.C. and Kusiak, A(eds). Handbook of design, manufacturing and Automation. John Wiley and sons, Inc New York Sgegheo, O. and Andersen, B (1999). Modeling the extended enterprise: a comparison of different modeling approaches, International Conference on Enterprise Modeling, Verdal. Stock, J.R. (1998) Development and Implementation of Reverse Logistics Programs. Council of Logistics Management. Thierry, M.; Salomon, M.; Nunnen, J. and Wassenhove, L. (1995) Strategic issues in Product Recovery Management. California Management review, 1995, Winter, 37(2), pp 114-135.
Wadhwa, S. and Madaan J (2004), Role of Quality Management (EFQM) Model for Reverse Logistics and GP: A KM View, ICQCC, International convention on quality control circles, Bangkok. Whitman, L., & Huff, B. (2001), On the use of enterprise models, International Journal of Flexible Manufacturing Systems, 13(2), pp 195–208.
660
Jitendra Madaan, Subhash Wadhwa & Mudit Verma* Indian Institute of Technology, New Delhi, India *MSIT, New Delhi, India Email: [email protected] Abstract: Direct sales channels, environmental legislation, and liberal returns policies have contributed to the growth of returns flows. However there is little empirical research related to the modeling of an integrated framework for returns. Till today, most of forward supply chain models focus only on physical transactions until the product reaches the customer’s end, leaving key decision variables implicit for some or all of the parties involved for return of product back from customer to the manufacturer. In this paper, we propose an alternative approach that explicitly addresses the extension of supply chain model for reverse logistics. As a matter of fact framework for decision modeling reverse logistics barely exists to support return. Therefore in order to fill this gap we discuss over the key decisions related with return handling and identifying quantitative models to support those decisions. Copyright © 2007 IFAC Key words: reverse logistics, decision making, and product recovery.
1. INTRODUCTION Conventional forward supply chain modeling usually considers a set of processes, driven by customer demand, that convey goods from suppliers through manufacturers and distributors to the end customers. But, this is not where the value of physical product terminates. Physical goods as well as their values do not simply get consumed fully once they have reached the customer. In order to capture this value we require a broadening of the supply chain perspective to include new processes, known as ‘reverse logistics’, and multiple interrelated usage cycles that are linked by specific market interfaces. Coordinating the successive activities after the product return is a key to maximizing the value generated. Therefore, many goods/products move beyond the conventional supply chain horizon, thus triggering additional business transactions: used products are sold on secondary markets; outdated products are upgraded to meet latest standards again; failed components are repaired to serve as spare parts; unsold stock is salvaged; reusable packaging is returned and refilled; used products are recycled into raw materials again. The set of processes that accommodate the flow of these goods ‘upstream’ in a conventional supply chain scheme, are known as ‘reverse logistics’. Today reverse logistics for returns have become endemic even in technologically advanced products, with rates as high as 20% in some sectors. Therefore, developing a comprehensive and cost-effective approach to handling returns is a daunting challenge that reaches well beyond the operational level. Thus, a well-developed reverse logistics and management plan can be a vital strategic asset. Further, supply chain management analyses have already revealed that this type of product returns can, in fact, be beneficial for both the manufacturer and the retailer. Thereby, the manufacturers benefit hinges on larger expected sales volume, competition marketing motives, direct economic motives, and concerns with the environment. Some of these are shown in figure 1.
655
Figure 1 Benefits Extracted From Reverse Logistics With the legislative measures tightening up, there are not many options left with the companies but to go for reverse logistics practices. According to a survey conducted by the Reverse Logistics Executive Council (RLEC), the average returns rate is 8.46% with individual expected return shown in table 1. Table 1: Expected Rate of Return (Survey by RLEC) Product category Return % in Year 2004 White goods 8% House Hold appliances 7% TV’s 8% Computers and 15% accessories Brown Goods 6% But looking across the entire manufacturing value chain, one finds return rates are as high as 20% or more in the year 2004 and these rates are expected to increase more in near the future. Therefore there is a need to achieve efficiency in resource consumption, reduction in waste and extraction of value from the return, when there in one out of five products is being returned. Thus in a closed loop system it will be necessary to develop new approaches for the systematic use and re-use of
resources in a broader sense. This paper presents a new approach for extending supply chain modeling framework for reverse logistics in order to capture opportunities for improved decision making. 2. MODELING FRAMEWORK FOR INTEGRATED REVERSE LOGISTICS SYSTEM Till today, most often companies have not even mapped processes related to them (Stock, 1998). Returned products are still treated on an ad hoc basis and regarded as waste. The lack of interest can partly be explained through the fact that managing product returns flows is complicated due to their interfunctional and inter-enterprise nature. Several methodologies like business process reengineering (BPR), enterprise modeling (Sgegheo & Andersen, 1999; Whitman & Huff, 2001), enterprise integration modeling (Petrie, 1992; Presley, 1997), and integrated enterprise modeling (IEM) have been suggested to improve enterprise performance through the designing and modeling of business processes. All these efforts concern themselves with the configuration of processes to achieve enterprise goals, with slightly different focuses in forward direction, but a lot hasn’t been done for modeling the framework with product return. RES (Reverse Enterprise System) requires enterprise and inter-enterprise relationships are viewed or modeled from return perspectives required which are to be fully described and understood in forward direction. Traditional supply chain modeling methodologies typically emphasize forward aspects keeping in view the chain ends as products reaches to customers end (Rogers & Tibben, 1999 & Thierry, 1995). However, here in this paper we look forward to model product return aspect from vastly different conceptual perspectives. Return process can be looked at from various perspectives depending on the kind of information required. Usually this will represent the activities comprising: when product came back into the chain; what kind of re-manufacturing functions are required to be done; who is performing these activities; how and why are they executed; when and where are these activities performed; and what data elements is required to be manipulated. Looking at the prior research, reverse logistics has multiple views or perspectives such as function (evaluation or screening, remanufacturing, repair, resell etc), information (related to product usage and return), behavior (uncertainty), organization (structure of organizing the return channel), resource (control the flow product in return chain)
656
Figure 2 Business Model for Strategic Reverse Enterprise System ,business and legislative rules, decisions, goals, and actors (or people) views for the return process. However, the existing methodologies that have been utilized for the modeling and integration of an enterprise do not specifically address the techniques for inter and intra reverse enterprise system modeling and integration. Therefore this paper, presents a generic model to support understanding, analysis, and design of product return processes aligned with information systems for inter-enterprise integration between enterprises as well as for a particular enterprise. We start with development of generic product recovery architecture for a reverse enterprise system as shown in figure 3.
Figure 3 Generic Product Recovery Architecture for a Reverse Enterprise System The development of product recovery architecture includes an integrated modeling framework and a suite of modeling techniques for enterprise modeling. The framework and methodology provide a systematic means to simultaneously achieve three kinds of integration: model integration of the different views, paradigm integration of the traditional and object oriented paradigms, and the integration of the reverse enterprise systems. The methodology here includes acceptance criteria which determine whether a material is suitable for specific uses or activities. These criteria are applied within the model to guide selection of materials for specific applications and their successive uses. 3. OPPORTUNITIES FOR IMPROVED DECISION MAKING IN REVERSE LOGISTICS Decision making process can be defined as a methodical approach of decision making that allows managers to handle problems where different alternatives and/or a certain degree of uncertainty are involved. Product return is well known for its uncertainty; therefore main objective of decision analysis in reverse logistics is to propose a model to assist decision makers in making better decisions. The decision problem for reverse logistics can be effectively solved just by using an appropriate decision making method. But, there are number of decision-making methods mentioned in the literatures [Stock, 1998]. The availability of a wide variety of decision-making
methods leads to another problem of choosing a suitable method. These decision-making methods differ widely in the purposes they serve, their ease of use and theoretical soundness, and the evaluations they yield. An intending user must thus consider the appropriateness of the method to the problem in terms of the value judgments it asks from the decision maker, the types of alternatives it can consider, and the forms of evaluations it yields. Although there have been some models developed for the evaluation of product after return options, they are either mono criterion- or bi-criteria based. Some of these efforts are focused only on some of the aspects or on one or two product return processes. The economics of design for re-cycling by using cost and benefit analysis method was given by Kroon & Vrijens [1995]. By this method the cost of each return option was first computed, followed by the calculation of the benefits of each of the options. The results of the cost and benefits calculations of each of the options were compared to determine the most profitable alternative. However, the focus of the work is on the product design with the consideration of the value alone, excluding the utilization stage. Furthermore, the return options considered are parts reuse/resale, product recycling, cannibalization, and remanufacture/ repair etc. while the basis of evaluation is limited to environmental and economic factors. Low and associates presented a number of mathematical models to assist designers in evaluating a number of return (reverse manufacturing) options of a product at the conception stage of the product development. The options being considered are recycling, remanufacturing, resale, repair, and disposal etc. The cost models evaluate the cost of each model as a fraction of the reverse manufacturing cost and consequently evaluate the trade-off between the options. Again the basis of evaluation is only financial and is directed at the product design. Moreover, authors have also reported a number of efforts on remanufacturing and disassembly [Fleischman, 2000, Giultinian, 1975, Krikke et al., 1999, Kroon & Vrijens, 1995 and Dowlatshahi, 2000]. Among them is the development of metrics for the assessment of remanufacturability of designs and for measuring ease of assembly, disassembly, testing, inspection, cleaning, and part replacement by Bras and associates. Here in this paper our model includes the both cost and utility based aspects of the product return process being studied. 4. DECISION MAKING APPROACH FOR REVERSE LOGISTICS This paper focuses on manager’s decision for reverse logistics process for product return. Problematic areas or challenges in the current product return practices were identified. These challenges were described in section 1 and 2. Based on the challenges described, the decision making process was studied to minimize the problems that could be encountered. The quality and quantity logistics decisions define the choice of reverse manufacturing process. Barry et al. [1993] and Carter& Ellram [1998] identified a number of process alternatives for product recovery. The five 657
notable ones among these reverse manufacturing techniques are repair and maintenance, refurbishing; remanufacturing; cannibalization, and reuse. The decision about each of these alternatives is initiated by functional level breakdown and flow block analysis. It is followed by the assessment of resource requirements for reverse logistics operation, and the data resulting from these are used to evaluate the suitability of the reverse manufacturing option. Figure 3 shows various reverse manufacturing options from firm to firm both for a particular product and for different products, a particular scenario can be chosen as per the availability of resources. 4.1 Evaluation and analysis of Decision for reverse manufacturing Various decision evaluation approaches and applications such as [Demmel, J.G. and Askin, R.G. (1986), Sanchez et. al (1994)] revealed that a product value function can be used to describe a relationship between a set of attributes of same dimension of value and the degree of utility corresponding to that attribute. Here the proposed value function can be applied to decision making process for reverse manufacturing function. The value or utility of each option can thus be calculated as a measure of preference for various values of a variable, having measured the relative strength of desirability that the decision maker has for those values. Suppose {α, α1, α2, α 3, α m} are the feasible reverse manufacturing function for the decision problem (represented by Figure 3), {β1, β2, βn} is a set of attributes, and δmn denotes a specific level of βn with regard to reverse manufacturing option αm. Here if we follow decision theory then certain preferential and independence conditions hold, then υ (δ11, δ12, δmn) has the form of a simple additive weighted utility value function: N
υ j (α ) = ∑ wiυ i (δ i ) i =l
Where υi(δi) = Value function for a attribute βi wi = weight-age given to the attribute βi υj(α) = The utility value of reverse manufacturing alternative αj on attributes ∑ {β1, β2, ..., βN} = The summation of the utility value at each of the attributes These generalised value functions can be rewritten for each reverse manufacturing alternative as follows: Repair process value function α 1 = w1[γ1(Dcost + O cost)]1 + w2 [γ2 (Sop + P Ch + Tcap)]1 + w3[γ3 (Rcon + Wrl + Wi)]1 + w4 [γ4 (Rc + Dr + Rr)]1 + w5[γ5 (Tr + Ts + To + T dly + Taux)]1 Refurbishing process value function α 2 = w1[γ1(Dcost + O cost)2 +w2 [γ2 (Sop + P Ch + Tcap)]2 + w3[γ3(Rcon + Wrl + Wi)]2 + w4[γ4 (Rc + Dr + Rr)]2 + w5[γ5 (Tr + Ts + To + T dly + Taux)]2 Remanufacturing process value function α 3 = w1[γ1(Dcost + O cost)3 +w2 [γ2 (Sop + P Ch + Tcap)]3 + w3[γ3(Rcon + Wrl + Wi)]3 + w4[γ4 (Rc + Dr + Rr)]3 + w5[γ5 (Tr + Ts + To + T dly + Taux)]3 Cannibalization process value function
α 4 = w1[γ1(Dcost + O cost)4 +w2 [γ2 (Sop + P Ch + Tcap)]4 + w3[γ3(Rcon + Wrl + Wi)]4 + w4[γ4 (Rc + Dr + Rr)]4 + w5[γ5 (Tr + Ts + To + T dly + Taux)]4 Reuse process value function α 5 = w1[γ1(Dcost + O cost)5 +w2 [γ2 (Sop + P Ch + Tcap)]5 + w3[γ3(Rcon + Wrl + Wi)]5 + w4[γ4 (Rc + Dr + Rr)]5 + w5[γ5 (Tr + Ts + To + T dly + Taux)]5
Where Dcost ---- Direct costs O cost ---- Overhead cost Sop ---- Product state after return P Ch ---- Process characteristics Tcap ---- Techno-capability Rcon ---- Resource consumption Wrl ---- Waste released Wi --- Waste impact Rc ---- Resources conserved Rr ---- Resources required to reprocess Dr ---- Demand of reprocessed product Tr ---- Time Reverse manufacturing option Ts ---- Set-up time To ---- Actual process operations time Taux ---- Auxiliary times T dly ---- Delay time γ 1 ---- Normalizing for cost attribute δ 1 γ 2 ---- Normalizing for technical attribute δ 2 γ 3 ---- Normalizing for environmental attribute δ 3 γ 4 ---- Normalizing for market attribute δ 4 γ 5 ---- Normalizing for time attribute δ 5
percentage – change-based analysis is applied by changing the estimates in increments of plus and minus ten percent and recompiling the results. Consequences of the uncertainty in the accuracy of data for evaluation, sensitivity analysis of how variation in data affects the overall performance of a reverse enterprise system and the effect of decision making process for various reverse manufacturing & logistics functions shown in fig 4 & fig 5.
4.3 Improved Decisions After assessing each reverse manufacturing alternative on the all attributes, the results have to be compared with the satisfaction of minimum standard on each of the attributes. The final selection of the alternative to be used for the extension of a particular product in a specific location can be based on three principles, namely: satisfying solution, maximization of expected utility, and preferred solution [Sanchez et al., 1994]. 4.4 Minimum acceptable solution The set of satisfying solutions consist of all processes that meet minimal requirements: l ⎡ l ⎤ aia = {ai ∑ wi vi ( f i ) ≥ ⎢∑ wi vi ( f i )⎥ } i =1 ⎣ i =1 ⎦ min
⎡
⎤
l
∑ w v ( f )⎥⎦ min ⎣
Where ⎢
4.2 Sensitivity Analysis
i i
i =1
i
= Minimum acceptable
performance.
Sensitivity analysis, which refers to the study of how important results changes with changes in estimates, is a “what-if” technique that looks at how a result will be changed if assumptions change or original estimates are not achieved. It is applicable in any analytical technique involving uncertainty in their underlying assumptions.
4.5 Maximum benefit solution This decision is for a decision-maker in favor of maximizing expected utility/benefit. In this case, recourse is not made to minimum satisfactory condition level with respect to any attribute. Thus this solution is purely based on compensatory method that permits tradeoffs between the attributes. Nj
amb = {ai max ∑ w f vi f ( f i ) j =1
⎡ Nj ⎤ ⎢∑ w j ⎥} ⎣ j =1 ⎦
4.6 Preferred solution These are the solutions which are both acceptable and benefit maximizing. The solutions are the one utilizing the integration of both compensatory and noncompensatory techniques, combining the advantages of the methods. This means, selecting the reverse manufacturing option so that:
Figure 4 Sensitivity of reverse manufacturing option
with respect to cost can either be Numerical sensitivity analysis
Nj
It is recognized as an aid for validating the model and for identifying model improvement possibilities [Sanchez et al., 1994]. Sensitivity analysis may be carried out numerically or by differentiation. Numerical sensitivity analysis can either be displayed as absolute amounts or as percentage changes from the base estimates or both. In this paper, the 658
aideal = {ai max ∑ w f vi f ( f i ) j =1
⎡ Nj ⎡ Nj ⎤ ⎢ ∑ w j ⎥} ≥ ⎢ ∑ w f vi f ( f i ) ⎢⎣ j =1 ⎣ j =1 ⎦
⎡ Nj ⎤ ⎤ ⎢∑ w j ⎥ ⎥ ⎣ j =1 ⎦ min ⎦⎥
Depending on the nature of the decision maker, represented by the three mentioned decision making principles earlier, substituting all the relevant values obtained from equations 3.3 - 3.7 and equations 2.65 -
2.71 into any of equations 3.8 - 3.10 results in an optimal reverse manufacturing process selection. 5.0 COMPUTER BASED IMPLEMENTATION OF DECISION MODEL The comprehensiveness of this model and the data requirements with the attendant calculations and analyses make the application of the methodology tedious without the use of a computer. A Computer application does not only quicken the implementation of the model but also facilitates easy and fine presentation of the implementation results. This model can be easily implemented on a computer by using any of the windows application programmes such as Visual Basic, Visual C++ and others. However, ARENA 7.0 simulation package is used in this work to develop the demonstrative computer implementation prototype. The prototype can later be upgraded to a decision support tool for reverse enterprise systems. This demonstrative computer prototype also supports the decision model in assessing other parameters like the life-extendibility of the retired product, marketability of the reprocessed product and the cost of adopting a specific process in extending the life of the retired product. Furthermore, it facilitates the evaluation of available facility’s suitability for the process and consequently for the chosen reprocessed product quality. The process time, and the conformity of the process to legislative requirement can equally be determined by using the computer application prototype. 5. CONCLUSION Despite all the research efforts thus far being put into resource recovery and reuse, none has been found to consider their decision-making aspects in a comprehensive manner (Wadhwa & Madaan, 2004). Most of the research paper focuses on the development of some aspect of reverse manufacturing techniques in the areas like product take-back logistics. Others focus on product design that facilitates resource recovery and reuse. The absence of a comprehensive decision making framework in the area of resource creates a void which this paper is attempting to fill. This paper proposed an integrated framework and model for improved decision making for reverse manufacturing option selection methodology, and a proposed scope for implementation of computer based implementation of the methodology. This paper, by establishing parameters needed for the evaluation of reverse manufacturing and logistics processes; by developing a suitable correlations for decision making in this domain; as well as by developing framework for setting minimum standard on major decision making parameters and demonstrating its applications, will be useful for decision makers. REFERENCES Barry, J.; Girard, G. and Perras, C. (1993) Logistics planning shifts into reverse. Journal of European Business, 5(1), 34–38. Carter, C.R. & Ellram, L.M. (1998) Reverse Logistics: A review of the literature & framework for future investigation, J. of Business Logistics, 19(1), 85102. 659
Demmel, J.G. and Askin, R.G. (1986). A multiple objective decision model for the evaluation of advanced manufacturing system technologies. J. Manuf. Systems 11(3), 133-152 Dowlatshahi, S. (2000) Developing a theory of Reverse Logistics. Interfaces, 30(3), 143-155. Fleischman M.; Krikke H.R.; Dekker R. and Flapper S.D.P. (2000) A characterization of logistics networks for product recovery. Omega, The International Journal of Management Science, 28(6),pp 653- 666. Giultinian, J.P. and Nwokoye, N.G. (1975) Developing distribution channels and systems in the emerging recycling industries. International Journal of Physical Distribution, 6(1), 28-38. Hillegersberg, J.; Zuidwijk, R.; Nunen, J. and Eijk, D. (2001) Supporting Return Flows in the Supply chain. Communications on the ACM, June, 44(6),pp 74-79. Krikke, H.R.; Harten, A. and Schuur, P.C. (1999) Business case Roteb: Recovery strategies for monitors. Computers and Industrial Engineering, 36(4), 739-757. Kroon L., Vrijens G. (1995). Returnable containers: an example of Reverse Logistics. International Journal of Physical Distribution and Logistics Management, 25(2), pp 56-68 Petrie, C. (Ed.), (1992). Introduction. First International Conference on Enterprise Integration Modeling, Austin, TX. Pohlen, T.L. and Farris II, M.T. (1992) Reverse Logistics in plastics recycling, International Journal of Physical Distribution and Logistics Management, 22(7), pp 35-47. Presley, A. R (1997). A representation method to support enterprise engineering. PhD Dissertation, Industrial and Manufacturing Systems Engineering, University of Texas at Arlington Logistics Executive Council. Reverse http://www.unr.edu/coba/logis Ritchie, L.; Burnes, B.; Whittle, P. and Hey, R. (2000) The benefits of Reverse Logistics: the case of Manchester Royal Infirmary Pharmacy. Supply Chain Management: an international Journal, 5(5), pp 226-233. Rogers, D.S. and Tibben-Lembke, R.S. (1999) Going backwards: R L trends & practices. Reverse Logistics Executive Council, Pittsburgh, P.A. Sanchez, J.M.; Priest,J.W. and Burnell, L.J. (1994). Design Decision Analysis and Expert systems in Concurrent Engineering. In: Dorf, R.C. and Kusiak, A(eds). Handbook of design, manufacturing and Automation. John Wiley and sons, Inc New York Sgegheo, O. and Andersen, B (1999). Modeling the extended enterprise: a comparison of different modeling approaches, International Conference on Enterprise Modeling, Verdal. Stock, J.R. (1998) Development and Implementation of Reverse Logistics Programs. Council of Logistics Management. Thierry, M.; Salomon, M.; Nunnen, J. and Wassenhove, L. (1995) Strategic issues in Product Recovery Management. California Management review, 1995, Winter, 37(2), pp 114-135.
Wadhwa, S. and Madaan J (2004), Role of Quality Management (EFQM) Model for Reverse Logistics and GP: A KM View, ICQCC, International convention on quality control circles, Bangkok. Whitman, L., & Huff, B. (2001), On the use of enterprise models, International Journal of Flexible Manufacturing Systems, 13(2), pp 195–208.
660