The performance of lot sizing algorithms incorporating back-orders is evaluated and compared with several of the traditional lot sizing rules for the multi-level environment. By means of simulation, the impact of a variety of demand streams, product structures and cost parameters on lot sizing is examined. The back-ordering algorithms offer greater flexibility with little or no decrement in performance when compared to their non back-order counterparts.
Material Requirements Planning (MRP) systems continue to serve as the focus of most companies' materials management process. Computer simulation is a tool that is widely used to analyze the performance of such systems. This paper considers the impact of the presence of both supply and demand uncertainties on the performance of MRP Systems. The work reported here is based on a simulation study involving a variety of supply and demand conditions. The study establishes the significance of product structure shape and variant, lot sizing rule, cost ratio and the uncertainty factors.
This paper considers the materials management process. Recently, the materials management process has undergone many changes. These changes raise new design, implementation and operations issues insofar as the materials management process is concerned. Thus, unraveling the materials management process is an ongoing requirement. While the materials management engineer draws upon a varied tool kit, perhaps the most powerful technique available is simulation modeling.
This paper introduces an easy alternative to the Wagner-Whitin backorder algorithm. The algorithm is described and illustrated with an example. Its performance is compared with several of the traditional lot sizing rules (lot for lot, economic order quantity, period order quantity, least unit cost, least total cost, part period algorithm, Silver-Meal algorithm and Wagner-Whitin algorithm) as well as the backorder versions of Wagner-Whitin and Economic Order Quantity. This comparison is conducted using four different demand streams and 25 combinations of carrying, ordering and shortage costs. An aggregate performance index has been defined to compare the lot sizing rules. It is concluded that this new algorithm is sufficiently robust and relatively easy to apply. The results of the experiments conducted highlight the more general nature of the backordering algorithms. By choosing the right combination of costs, backordering can be suppressed and traditional lot sizing behavior can be imitated.
This paper investigates the impact of planned backordering in the context of demand uncertainty for a multi-level product structure environment operating under a rolling horizon. To this end, the performance of lot sizing algorithms incorporating backorders is evaluated for such an environment by means of simulation modeling. Given the presence of demand uncertainty, the results suggest that backordering algorithms can sometimes improve performance.
The influence of uncertainty on the performance of Material Requirements Planning (MRP) systems has been the focus of several studies. All of these studies have incorporated non back-order lot sizing rules. However, attention has been recently directed to back-order lot sizing rules. This paper considers the impact of the presence of both supply and demand uncertainties on the performance of MRP Systems in an environment where back-orders are planned. The work reported here is based on a simulation study involving a variety of supply and demand conditions. The study establishes the significance of product structure shape and variant, lot sizing rule, cost ratio and the uncertainty factors.
A knowledge based simulation model for combined Just-in-Time/ Material Requirements Planning systems is presented. The model written in PROLOG includes two expert systems, one for kanban logic and the other for job shop control. The model is applied to an electronics assembly plant. Factors related to demand variability, vendor lead time variability and buffer sizing are examined. Knowledge based simulation modeling is found to be an effective approach for examining such combined systems in a real world environment.
In this paper, the notion of duality and a generalized methodology to identify the dual for an Erlang distribution based truncated steady state queueing process is suggested. This procedure provides insight into truncated queueing processes. At first glance, the relationship between the primal problem and the dual problem may not be intuitive. Several such relationships are identified. For example, the dual of the M/Es/C/C queueing process is shown to be the M/M/1 machine interference problem with C machines and the solution to one of these provides the solution to the other. An example is presented.
The effects of the presence of both demand and lead time uncertainties on Material Requirements Planning systems performance are investigated. A simulation model of a multi-level product environment is employed to evaluate the influence of these combined uncertainties in a rolling planning horizon. Their interaction with product structure variables, cost ratios and lot sizing rules is also explored. Finally, by means of statistical analysis of the simulation output, hypotheses relating to various variables and their interactions are formed and tested. The results highlight the significant differences between uncertain and deterministic environments.
This paper addresses the increasingly prevalent phenomenon of customer integrated production systems. In particular, it examines the operation of systems where customer integration has been enabled by the implementation of EDI and where supply chain uncertainty exists. A simulation model of such a system was employed to analyze four different production systems operating under different batch procedures and cost scenarios. Numerous experiments were conducted and the results obtained were analyzed. Several conclusions relating to customer behavior pattern and supply chain uncertainty were drawn from this analysis.
In this paper, the interrelationship between a queueing system with finite capacity, balking and reneging and a machine repair problem with warm spares are formalized. It is shown that the recently reported relationship between an Erlang loss system and a finite source queuing system is a trivial consequence of the more general results presented here. Relationships between arrival point probabilities and measures of performance for such related queueing systems are also derived.
In this paper, the concepts of complementarity and equivalence in finite source queueing models with spares are formalized. The relationship between two finite source queueing models with spares is established. It is shown that all the well known relationships between queueing systems involving finite capacity and finite source are special cases of the more general results presented here. Relationships between arrival point probabilities and measures of performance for such related queueing systems are also derived.
We present an algorithm for scheduling the disassembly of discrete parts products characterized by a well defined product structure. As opposed to MRP where the demand occurs at the end item level, the demand in the disassembly case is motivated by the component level of the product structure. Even though the objective in the disassembly case is the reverse of that of MRP, the algorithm itself is not the reverse of the MRP algorithm. It is considerably more complicated. An example is presented and its implementation in a spreadsheet is described.
The performance of lot sizing algorithms incorporating back-orders is evaluated and compared with several of the traditional lot sizing rules. This comparison is undertaken for a multi-level product structure environment operating under a rolling horizon in the presence of lead time uncertainty. By means of simulation, the impact of a variety of lead time scenarios, product structures and cost parameters on lot sizing is examined. It is shown that unlike in the deterministic environment, the back-ordering algorithms can yield superior performance in an uncertain environment.
In this paper, numerous complementary and equivalence relationships between various Erlang based queueing systems are established. It is shown that some of the recently reported results are a trivial consequence of the more general results presented here. The implications of these relationships are discussed. It is concluded that these relationships, not only facilitate solutions to many Erlang based queueing systems, but also provide greater insight into such systems. An example is also provided.
The establishment of disassembly plants and the creation of product designs which specifically facilitate disassembly are enabling manufacturers to carry out item segregation. Item segregation is defined as the separation from an assembly of a part or a group of parts by following a reverse assembly process. Once segregated, the items can be reused, recycled or discarded. However, there are operational problems associated with item segregation. Foremost amongst these are the lack of planning and scheduling mechanisms, difficulty in coping with reverse flow of materials, and item explosion. Despite the economic and environmental benefits of disassembly, researchers and practitioners are lagging behind in developing methodologies to address the operations and production planning and control issues associated with item segregation. This paper is aimed at addressing these issues.
In this paper, some very useful and applicable ideas are presented to facilitate solving complex problems in Queueing Theory. It is demonstrated how a spreadsheet can be used to solve problems which many practitioners find very intimidating. To this end an algorithm is presented which is particularly designed for easy implementation in a spreadsheet. A template is provided illustrating the implementation of the algorithm. The use of the template is demonstrated in various queueing applications.
A closed loop type manufacturing flow line is considered where the fastest station can be used for additional work other than the regular job flow. The exact throughput of the system is derived and the moments of the cycle time distribution are approximated. Upper and lower bounds are given for the mean and variance of the cycle time.
This paper presents an approximation methodology for the analysis of a flexible manufacturing system (FMS). The FMS considered consists of split and merge configuration. Furthermore, the machines are subject to failure. The system is modeled using closed queueing networks. The approximation methodology first decomposes the system into individual M/M/1/K stations. Next, it aggregates the split and merge nodes into a unified node. Finally, the resulting cyclic network is used to find the throughput of the system. The machine breakdowns are modeled using queueing models with vacations. Two examples are presented to illustrate the behavior of small flexible manufacturing systems with split and merge configuration and machine failures.
This paper reports a model of a real time manufacturing system. Within this system the ongoing operations of the shop floor are monitored and up to the minute information is maintained. Such a manufacturing system incorporates information as soon as it becomes available and assembles and processes items only if the required parts from the preceding stages are available. In addition, the manufacturing system is interfaced with the customer by means of electronic data interchange (EDI). This allows instant communication between the manufacturing enterprise and the customer. The model reported here incorporates these features. Thus, the real life impact of uncertainties as they relate to an integrated shop floor control environment is captured by the model. Several hypotheses relating to various variables and their interactions are formed and tested. The differences between the deterministic and the real time environment are highlighted.
This paper presents stochastic analysis of repairable systems involving primary as well as secondary failures. To this end, two models are considered. The first model represents a system with two identical warm standbys. The failure rates of units and the system are constant and independent while the repairs times are arbitrarily distributed. The second system modeled consists of three repairable regions. The system operates normally if all three regions are operating, otherwise it operates at a derated level unless all three regions fail. The failure rates and repair times of the regions are constant and independent. The first model is analyzed using the supplemental variable technique while the second model is analyzed using the regenerative point technique in the Markov renewal process. Various expressions including system availability, system reliability and mean time to system failure are obtained.
We investigated the effect of supplier and process uncertainty in a real time MRP environment. The environment modeled is supported by an "on line" data monitoring and data collection system such that the activities of the shop floor and the receiving area are incorporated within the MRP database as they occur. The purpose of the model is to evaluate the operation of an integrated shop floor control/MRP system in a multi-level product structure environment which operates on a rolling horizon basis. The interaction of supply/process uncertainty with product structure and choice of lot sizing rule is examined for various cost factors. The presence of uncertainty is found to greatly complicate the operation of the MRP system. In addition, supply/process uncertainty combines with product structure and the choice of lot sizing rule, to impact the cost performance of the system. The significance of these interactions was confirmed by the formulation and testing of several hypotheses. Based on these results, some practical interpretations are also presented.
This paper describes the implementation of JIT in a small manufacturing company and the benefits that resulted for the company's operations. Preliminary analysis identified various problems in the existing manufacturing operations. The pre-implementation and post-implementation conditions of the company are detailed. The achievements of the JIT implementation included a reduction in material traversing, reduced lead times and inventories leading to an overall reduction in the cost of manufacturing. A smooth flow of material from the raw material stage through the finished product stage was established. Three separate product lines were combined into a flexible manufacturing assembly line. With the adoption of a holistic approach to JIT implementation, it was found that even a small company can make significant strides towards world class manufacturing status. The experience gained by the company can encourage and benefit other small companies to embrace the JIT approach.
It is well known that the performance of the JIT production system is optimum in a deterministic environment. However, real-world situations contain uncertainties in processing times, with which traditional JIT does not deal well. In this paper, we present a newly developed Kanban system that systematically manipulates the number of Kanbans to cope with the discrepancies introduced by the uncertainty in processing times. We illustrate that the performance of this new system is superior to the traditional JIT system in such an environment.
We consider a single stage, make-to-order job shop with unit demands, units production and a finite capacity. The production can be switched on or off depending on the level of orders waiting to be processed. Under a threshold type operating policy, optimal production switchover levels. that minimize a cost function including holding, operating and switching costs are found.
This paper presents the basic structure for two disassembly scheduling algorithms applied to a single product structure. The first algorithm addresses the case when all items in the product structure are unique, The second algorithm accounts for common items.
In this paper, the concepts of complementarity and equivalence between an M/M/c/K queueing model with state dependent balking and reneging and a machine interference problem with warm standbys are formalized. The relationship provides insight into these queueing systems. Through a series of corollaries, relationships between various queueing systems are derived. It is shown that a recently reported relationship between Erlang loss system and a finite source queueing system is a trivial consequence of the more general results presented here. New results involving the arrival point probabilities and measures of performance for these two queueing systems are also presented. An example is also provided.
In this paper, the interrelationship between the F-Policy problem and the truncated N-Policy problem is considered. The F-Policy problem deals with the issue of controlling arrivals to a queueing system. The N-Policy problem involves a queueing system in which the server leaves the system (for alternative jobs) when it becomes empty and returns when the queue builds up to a predetermined level. The steady state probability distribution and expressions for the measures of effectiveness for these two systems are obtained. The complementarity relationships between these two queueing systems are established. These relationships provide insight into these queueing problems and facilitate the solution to any one of the problems given the solution to the other. Through a series of propositions, other relationships between these queueing systems are also established.
The introduction of finite buffers in a queueing network destroys the product form behavior of the network and makes its analysis very difficult. Many researchers have proposed various approximation techniques to overcome this situation. However, in almost all cases, the researchers have assumed that the servers in the network are reliable and do not breakdown. The addition of unreliable servers, complicates the problem even further. We address this situation and present a newly developed expansion algorithm for the analysis of tandem networks with finite buffers and unreliable servers and "Block After Service" (BAS) blocking mechanism. We compare its performance with results obtained from simulation models. The results show that the algorithm is very efficient and robust.
This paper presents an overview of SCPN theory, a set of three generic SCPN submodels, a behavior-preserving methodology for combining those submodels into a larger model, and a combined model. The three submodels include a machine with failures, repairs, and limited I/O buffers, and two types of conveyor submodels - a simple conveyor segment and a conveyor segment at a machine.
In this paper, we present the results of a newly developed Kanban system that systematically manipulates the number of Kanbans to cope with the interruptions introduced by machine breakdowns. We illustrate that the performance of this new system is superior to the traditional Kanban system in such an environment.
In this paper, the machine interference problem is considered under N-policy service with a startup time (or setup time). The steady state distribution of the number of machines in the system is sought. Closed form solution for such a system is not attainable. However, we observe that it is possible to calculate the steady state probabilities recursively and we derive closed form expressions for them. Also derived are the expressions for the measures of effectiveness of these types of models. Several special cases are considered. It is shown that one of the special cases of the model can easily be extended to incorporate various vacation models. Closed form expressions for those are provided. An efficient and generalized algorithm is presented to find the steady state distribution of the number of machines in the system as well as the measures of effectiveness. An example and some sensitivity results are also presented.
In this paper, we present a newly developed algorithm which dynamically adjusts the number of Kanbans in a Just-In-Time (JIT) system in order to offset the blocking and starvation caused by the variations in processing times and demand during a production cycle. We provide steps of the algorithm and describe the methodology to model such systems. An example is considered and its results are discussed.
The performance of two lot sizing algorithms incorporating back-orders is evaluated and compared with several traditional lot sizing rules. This comparison is undertaken for a multi-level product structure environment operating under a rolling horizon in the presence of both demand and lead time uncertainties. By means of simulation and statistical analysis of the simulation output, the impact on system performance, of a variety of demand and lead time scenarios, product structures, cost parameters and lot sizing, is evaluated. The conclusions drawn emphasize important differences in the influence of the factors on system performance given the presence or absence of uncertainty.
With the increasing use of lighter materials to enhance fuel economy, the steel-dominated materials content in automobiles is changing to include a greater fraction of polymers and aluminum. This change may substantially impact the end-of-life (EOL) automobile disposal process. In light of impending regulations, various alternatives for component and material disposal are under investigation. Tradeoffs between technological and economic feasibility and the degree of environmental detriment must be optimized. This paper presents a preliminary application of a newly developed decision tool, called physical programming, to assess the alternative EOL disposal strategies of future vehicles. Physical programming addresses problems involving multiple objectives and constraints and allows the decision maker to express his/her value-system in a realistic manner for each objective of interest.
We consider the GI/M/1/K queue where the server takes exponentially distributed vacations when there are no customers left to serve in the queue. We obtain the queue length distribution at arrival epochs and random epochs for the multiple vacation case. We present heuristic algorithms to compute the blocking probability for this system. Several numerical examples are presented to analyze the behavior of the blocking probability and to test the performance of the heuristics.
In this paper, we use stochastic, colored Petri nets (SCPNs) to model a JIT system and simulate system performance under two different kanban control policies: traditional kanban system (TKS) and flexible kanban system (FKS). In TKS, the number of kanbans is fixed throughout the production cycle; the control problem is to determine the optimal number of kanbans. In FKS, the number of kanbans is systematically manipulated during the production cycle to improve system performance; the control problem is to determine when and by how much to manipulate the kanbans. The performance of the two systems is compared and FKS is shown to be superior for exponential processing times.
Petri nets (PNs) have recently emerged as a promising approach for modeling flexible and automated manufacturing systems. PNs are a graphical and mathematical modeling technique that is useful for modeling concurrent, asynchronous, distributed, parallel, nondeterministic, and stochastic systems, as attested by a steady stream of papers which appeared throughout the 1980s. Since 1989, there has been a explosion of interest in using PNs to model, simulate, and analyze manufacturing systems. We present an overview of PN theory. We then present a schema for categorizing PN models of manufacturing systems, followed by a survey of PN models of flow lines, automatic transfer lines, job shops, flexible manufacturing systems, and assembly systems. This discussion represents the most comprehensive survey of applications of PNs to manufacturing through 1994. Finally, we present a summary and some conclusions.
In this paper, we present an algorithm to obtain a disassembling scheme for the complex problem involving multiple product structures where the parts/materials have multiple occurrences. In particular, the algorithm determines the quantity and operations schedule of disassembly for all product structures (including the ordering of the roots and the disassembly schedule for the roots and the subassemblies) in order to fulfill the demand for the various parts. An overview of the algorithm is presented together with the results of an example to illustrate the workings of the algorithm.
With incoming recycling regulations, resource conservation needs and an increased awareness of the state of the environment by both the consumer and the producer, a fundamental reappraisal of the traditional manufacturing paradigm has been emerging. The manufacturers are under tremendous pressure to dispose of products in an environmentally responsible manner. To this end many companies are establishing disassembly plants and developing product designs which specifically facilitate disassembly. Once disassembled, the items can be reused, recycled or discarded. This paper provides an overview of the research in the area of disassembly of products.
Disassembly has been emerging as an important potential response to recent environmental and recycling regulations as well as resource conservation needs. One can identify two distinct approaches to tackle the disassembly problem, viz., design for disassembly (DFD) and planning for disassembly (PFD). This paper focuses on the PFD approach. The evolution of a disassembly plan may require some heuristics and domain specific knowledge. In addition, disassembly planners may have a particular style to solve disassembly problems. Due to these facts and the nature of the PFD itself (open ended and iterative), the paper proposes the use of Analogical Problem Solving (APS) as an approach to assist planners to solve PFD problems. APS is based on the sensible notion that problem solving can be assisted by the reuse of solutions to similar problems encountered in the past. The framework of applying APS to PFD, the merits of this approach, and the issues relating to its use for PFD are discussed.
Surface Mount Technology (SMT) is a popular method of Printed Circuit Board (PCB) assembly in which high speed automated assembly machines are capable of placing in excess of 40,000 components per hour. In order to achieve these impressive assembly rates, complex placement machines must be programmed efficiently. Intrinsic to the configuration of these machines and the assembly process are some well established Operations Research problems. This paper addresses the feeder sequencing problem for an assembly machine with a sliding feeder rack in conjunction with a X-Y positioning table and a turret placement mechanism. This problem is a Quadratic Assignment Problem, and is proven to be NP-complete. Two different heuristic methods are proposed, each with unique characteristics that have the potential to be beneficial to an assembly operation dependent upon the restrictions of the planning task. One method is to assign feeder slots based on the transition between component types that naturally occur in the board placement path. The second method begins with an initial slot assignment and identifies exchanges between pairs of slots that generate improvements in the objective function. Minimizing the feeder travel distance over an assembly is the goal of each heuristic. A comparative analysis between the two heuristics is performed. Examples are presented and the attributes of each method are discussed. Arguments are presented to support "near-optimal" solutions to the problem. Given the complexities of the system, proper planning of the assembly process can take advantage of the independent control of each mechanism to create a natural relaxation of specific constraints.
This paper proposes a methodology for efficient process planning of concurrent machines in electronics assembly. The particular machine type under consideration is the High Speed Chip Shooter (HSCS) for surface mount assembly. Currently most surface mount assembly operations are modeled in a cyclic manner. The model that is proposed herein is an asynchronous (acyclic) model that augments the benefits of the unique features of this assembly machine. A heuristic algorithm is developed, referred to as the Acyclic Assembly Time (AAT) algorithm, which is based on the asynchronous model. The algorithm is thoroughly tested with orthogonal arrays, compared against previously published problems, and applied to a real life example. The AAT algorithm produced excellent results throughout the test process. By increasing the utilization of the individual mechanisms, the efficiency of the overall system naturally improves and the ultimate goal of reducing the total assembly time is achieved.
Optimization problems can be partitioned into two categories hereby called blind optimization and physical optimization. In blind optimization the analyst has no knowledge of the physical meaning of the problem at hand, or the nature of its anticipated solution. In physical optimization the analyst (in this case a decision maker) has substantive knowledge and often clear objectives, regarding aspects of the problem at hand, that can be articulated in physically meaningful terms (e.g. a Return On Investment of 10% is desirable; 8%, tolerable; and 5%, unacceptable). Nearly all operational research or engineering design problems belong to the latter category. A new optimization philosophy, known as physical programming, that explicitly addresses the physical optimization problem has recently been proposed, with successful initial applications in the domain of engineering design. This paper develops the concept of linear physical programming for operational research applications, by addressing the distinct issues related to complex linear systems.
This paper reviews the problems that many electronics manufacturers are facing in a society of rules and regulations that are becoming increasingly environmentally conscious. The effect of electronics assembly, disassembly, and disposal on the environment is reviewed and the potential hazards of continuing the present trends in electronics parts disposal is discussed. The paper contains a comprehensive survey of previous work related to environmentally conscious manufacturing practices, recycling, and the complexities of disassembly in the electronics industry. Interest in this area has intensified in the recent years due to an increased awareness of the problem in a world of high technology, where electronic products dominate. Industrial applications of recycling programs are presented and existing methodologies and evaluation systems are discussed. In order to promote and support this new environmental ethic in electronics assembly and disassembly, the need for improved methods of electronics reuse, minimization of life-cycle scrap, development of planning tools, and an increase in research activity in this area is also highlighted.
In a real life environment, the Just-In-Time (JIT) system is subjected to various types of uncertainties such as stochastic processing times and variable demand. Since, JIT was only meant to operate in a deterministic environment, its performance is seriously affected by variations in processing times and demand. In this paper, a newly developed Kanban system is presented which uses an algorithm to dynamically and systematically manipulate the number of Kanbans in order to offset the blocking and starvation caused by the said uncertainties during a production cycle. The new system is termed a Flexible Kanban System (FKS). The steps of the algorithm are detailed and the effectiveness of FKS is demonstrated using an example model. For the example model, the solution procedure, results and a discussion are presented.
We consider the machine interference problem with warm spares in which the server takes a vacation of random duration every time the repair facility becomes empty. We address the cases of multiple vacations, single vacation and hybrid multiple/single vacation schemes with exhaustive service. Vacation models that have been studied in the past generally assume infinite population. In this paper, we provide new, transform free, closed form expressions for the probability distribution of the number of machines in the repair facility and the performance measures for machine interference problem with warm spares and server vacations. By adjusting the parameters, closed form expressions for machine interference problem with cold spares, machine interference problem without spares, finite capacity and infinite capacity vacation models can also be derived. An extremely powerful and efficient algorithm is presented to determine the steady state probability distribution of the number of customers in the system as well as the performance measures for all these models. An example is considered and some insight is also provided.
This paper addresses the issue of parts and materials commonality when scheduling disassembly. In a disassembly environment, inventory management is complex due to the presence of multiple demand sources at the component level of the product structure. Commonality introduces a new layer of complexity by creating alternative procurement sources for the common component items. A novel scheduling algorithm is presented, followed by an example.
With recycling regulations, resource conservation needs and an increased awareness of the state of the environment by both the consumer and the producer, many companies are establishing disassembly plants and developing product designs which specifically facilitate disassembly. Once disassembled, the items can be reused, recycled or discarded. One can identify two distinct aspects of the disassembly problem, viz., design for disassembly (DFD) and planning for disassembly (PFD). The goal of DFD is to design products that are easy to disassemble. On the other hand, the objective of PFD is to identify efficient sequences to disassemble products. This paper focuses on the PFD aspect of disassembly. Due to the fact that there could be many ways to disassemble a given product, the PFD knowledge is accumulated by experience. Such knowledge is valuable and should be captured, saved and re-used to solve similar problems that arise in the future. In this paper, we propose Case-Based Reasoning (CBR), as an approach, to solve PFD problems. CBR is based on the fundamental principle that problem solving can benefit from solutions to past problems that have been attempted. The technique and issues related to the application of CBR to PFD are presented.
We consider finite buffered queues with service or arrival control. In the case of service control, service may be stopped and restarted depending on the queue length. In the case of arrival control, the arrival stream can be turned off and on or arrivals may be rejected depending on the queue length. We give duality relations for various systems with arrival and service control that enables us to relate their stationary queue length distributions. We use physical coupling arguments which imply the stochastic coupling necessary to relate the queue lengths. We also discuss special cases for which queue length relationships can be obtained by analyzing the underlying Markov process. Two examples are provided to demonstrate the application of the duality property. The first example is a case where the existing queue length distribution for a given model can be used to obtain the queue length distribution of another model. In the second example, we obtain the previously unknown queue length distributions for two related models at once.
In this paper, we address the problem of scheduling the disassembly of discrete parts products characterized by well defined product structures. We allow for the existence of multiple product structures as well as the existence of common parts and/or materials which makes the problem very complex. To this end, we present two companion algorithms which can be applied to obtain a disassembling scheme for such problems. Specifically, the algorithms determine the quantity and operations schedule of disassembly for all product structures (including the ordering of the roots and the disassembly schedule for the roots and the subassemblies) in order to fulfill the demand for the various parts. An example is presented to illustrate the use of the algorithms.
This paper proposes a sequencing approach to develop an efficient feasible path for the printed circuit board assembly process. Determining the component placement sequence, also referred to as the placement path, is an NP-complete problem that best resembles a Traveling Salesman Problem (TSP) for which a heuristic is developed. The heuristic approach is tested against a previously published subproblem as well as a real-life working board configuration. This heuristic is intended to provide a good, feasible component placement sequence for the assembly of a batch of printed circuit boards with an assembly type configuration consisting of a moveable X-Y positioning table and a tape-and-reel sliding feeder rack. Even with high speed assembly machines placing in excess of 40,000 components per hour (cph), process improvements are possible by increasing the efficiency of the planned placement sequence. This heuristic is developed to identify an improved component placement sequence in a reasonable computational time to allow for future implementation of the methodology in applied situations where time constraints are unavoidable.
We consider finite buffered queues where the arrival process is controlled by shutting down and restarting the arrival stream. In the absence of holding costs for items in the queue, the optimal (s,S) policy can be characterized by relating the arrival control problem to a corresponding service control problem. With the inclusion of holding costs however, this characterization is not valid and efficient numerical computation of the queue length probability distribution is necessary. We perform this computation by using a duality property which relates queue lengths in the controlled arrival system to a controlled service system. Numerical results which analyze the effect of setup and holding costs and the variability of the arrival process on the performance of the system are included.
Alternative disposal strategies for vehicle design with varying relative proportions of materials, are explored using goal programming to analyze the tradeoffs between technological, economic, and environment factors. Two vehicle designs - one based on a steel unibody and the other more intensively designed with polymer materials - were selected for investigation. The preliminary results indicate that if properly controlled, the current automobile recycling infrastructure in the U.S. can remain economically viable while improving with respect to environmental considerations.
Environmental awareness regarding resource use and emissions over the life cycle of the automobile has heightened the concerns for end-of-life (EOL) vehicle disposal. With increasing use of lighter materials to enhance fuel economy, the steel-dominated content of automobiles is changing to include a greater fraction of polymers. This change may substantially impact vehicle disposal. In light of impending regulations, various alternatives for remanufacturing and reuse of components and material disposal are under investigation. For example, if shredder operations are used to reclaim metallic materials, then the extent of disassembly will significantly impact profitability as well as the environment. Therefore tradeoffs between technological and economic feasibility, and the degree of environmental detriment must be identified for disposal scenarios of interest. Using goal programming, changes to the current US vehicle recycling infrastructure are explored for their effects on dismantler and shredder profitabilities. To investigate the effect of lightweighting on the profitability of the recycling infrastructure, two specific vehicle designs are compared: a steel unibody and a polymer intensive vehicle. Other scenarios examine the outcomes for mandating removal of polymer materials during disassembly, and for increasing the disposal cost of scrap polymer to that of hazardous waste. Goal programming addresses multi-objective problems involving linear multiple criteria and linear constraints, and allows the exploration of the vehicle recycling infrastructure profitability for prescribed target profits under varying conditions. These results indicate that if properly controlled, the current automobile recycling infrastructure in the US can remain economically viable while improving with respect to environmental considerations. Alternatively, implementation of certain policies that reduce profitability could cause disastrous consequences, resulting in the economic collapse of the infrastructure.
Disassembly is a systematic process that allows reusable, non-recyclable, and hazardous subassemblies to be selectively separated from recyclable ones. In this paper, we present a methodology to evaluate different disassembly strategies so that the best one could be chosen. Since the identification of all possible disassembly sequences of complex products is not an easy task, we also propose a disassembly sequence generation heuristic which gives a near optimum disassembly sequence for a product. The application of the methodology is illustrated by considering an IBM PS/2 Model 30 computer base.
The majority of modern day products contain thousands of parts and many different technologies. Many parts are reusable and some even possess a higher reliability rating than their new counterparts. In order to maintain the integrity of reusable parts, disassembly process planning has to make sure that the identified parts are retrieved properly. Planning for a disassembly process, as the number of parts increases, becomes more and more complex. In a product with 10 parts, there are potentially 10! (3,628,800) possible disassembly sequences. Thus, proper planning is necessary in order to identify the best plan. This research proposes a methodology to generate a disassembly process plan according to the product's modularity.
Environmental awareness regarding resource use and emissions over the life cycle of the automobile has recently grown and has heightened concerns for end-of-life (EOL) vehicle disposal. With increasing use of lighter materials to enhance fuel economy, the steel dominated materials content in vehicles is changing to include greater fractions of polymers and aluminum. This change may substantially impact vehicle disposal. In light of impending regulations, various alternatives for remanufacturing and reuse of components and material disposal are under investigation. For example, if shredder operations are used for reclaiming metallic materials for disposal, then the extent of disassembly significantly impacts the process profitability as well as the environment. Tradeoffs among technological and economic feasibility, and the degree of environmental detriment must be determined for each alternative. Using goal programming techniques and a model of the automobile recycling infrastructure, materials streams and process profitabilites are tracked for different processing scenarios. Optimal quantities of disassembled materials prior to shredding are determined which would maintain a prefered level of profitability for both the disassembler and the shredder. The three vehicle designs are considered which focus on materials in a specific class, i.e., plain carbon steels in a standard unibody vehicle, polymer composites based on body panel substitution in a unibody and aluminum alloys based on a space frame design. For each of these vehicle designs, the system model is used to generate information regarding the profitability of the recycling infrastructure. The steel unibody design is used as a base case to simulate current recycling infrastructure conditions. Various scenarios are then run to demonstrate the consequences for disassembler and shredder profitabilities with changes in: i) decreasing the ferrous content in the vehicle design, ii) increasing the quantity of polymer materials removed from the EOL vehicle during disassembly, and iii) increasing disposal costs for automobile shredder residue. The results indicate that the current automobile recycling infrastructure in the U.S. can remain economically viable while improving with respect to environmental considerations.
Disassembly is a key element for retrieving the desired subassemblies and/or parts from a product. However, determining an efficient disassembly sequence plan (DSP) is an NP-complete problem. In this paper, we propose a methodology to generate a near optimum DSP for a product. The methodology is illustrated using an example.
This paper presents a two-stage model that provides a unique solution for planning components recovery from products with components commonality. The objective of the model is to determine the number and type of products to disassemble, to fulfill the demand of various components, in order to minimize disassembly and disposal costs.
Disassembly is one of the proposed solutions to today's increased environmental problem of large-scale disposal of manufactured products. Disassembly process brings with it a lot of unresolved material control issues. In this paper we illustrate the implementation of the recently developed Flexible Kanban System to cope with the uncertainties that are unique to the disassembly system.
This paper presents a procedure to disassemble electronic products with multiple subassembly modules. First, a partial schedule for each subassembly is obtained. The next step modifies the partial schedule in order to minimize the machine idle time at the retrieval process and, thus, the resulting makespan of the whole process. The procedure offers an optimal process makespan according to the sequence in which the batch of products pass through the disassembly and recovery processes. Special emphasis is placed on applying variant process planning methodology for disassembly and retrieval.
Recycling and remanufacturing are important forms of product/material recovery which involve product disassembly to retrieve the desired parts and/or subassemblies. Disassembly is a systematic method for separating a product into its constituent parts, components or other groupings. Efficient disassembly requires development of disassembly sequence plans (DSPs). Generating DSPs describing the sequence of parts during disassembly is not a trivial problem since DSP generation is described to be NP-complete. Further complicating matters is the presence of a high degree of uncertainty due to upgrading/downgrading of the product during its use by the customers and defects occurring either when in use or during disassembly. In this paper, we address the uncertainty related difficulties in disassembly sequence planning. To this end, we present a methodology to develop a framework for dealing with uncertainty in DSP implementation and demonstrate it using a simple example.
Recycling and remanufacturing involve product disassembly to retrieve the desired parts and/or subassemblies. Disassembly is a systematic method for separating a product into its constituent parts, components, or other groupings. Disassembly process planning is critical in minimizing the amount of resources (e.g., time and money) invested in disassembly and maximizing the level of automation of the disassembly process and the quality of the parts (or materials) recovered. We propose an algorithm which automatically generates a disassembly Petri net (DPN) from a geometrically-based precedence matrix. The resulting DPN can be analyzed to generate all feasible disassembly process plans (DPPs), and cost functions can be used to determine the optimal DPP; alternatively, heuristic methods may be used to generate near-optimal DPPs.
This paper presents a technique to control the material flow in a disassembly environment using the Flexible Kanban System (FKS). The implementation and effectiveness of the FKS is demonstrated using a case example.
This paper considers a two-echelon inventory system with return flows, where demand and return rates are mutually independent. An open queueing network with finite buffers is used to model the system. The model is analyzed using the expansion methodology.
This paper presents a two-stage disassembly and retrieval system that provides a unique solution for planning selective component recovery from products. The solution approach uses process sequencing technique to schedule the operations in both stages to optimize the operational makespan.
We generate a disassembly PN (DPN) from a disassembly precedence matrix. The resulting DPN can be analyzed using the reachability tree method to generate all feasible disassembly process plans (DPPs), and cost functions can be used to determine the optimal DPP. Since generating the reachability tree is NP-complete, we develop a heuristic algorithm to limit the size of the reachability tree. The algorithm employs multi-hypothesis search to dynamically explore the v likeliest lowest cost branches of the tree, in order to identify near-optimal DPPs. The cost function incorporates tool changes, changes in direction movement, and individual part characteristics (e.g., hazardous). An example is used to illustrate the procedure.
This paper develops a methodology for the analysis of tandem manufacturing systems where a machine takes a vacation (i.e. becomes unavailable for some reason such as processing secondary jobs or being repaired) of random duration every time the corresponding station becomes empty. The system is modeled using a queueing network in which decomposition, isolation and expansion methodologies are used to calculate the throughput. The methodology was rigorously tested by designing experiments using orthogonal arrays to cover a large experimental region. The results are compared with benchmark results obtained through simulation. The differences between the two results are investigated using t-tests. The methodology developed proved to be robust and very accurate.
The Just-In-Time (JIT) system is designed to operate in an ideal environment such as constant processing times, smooth and stable demand and uninterrupted processing. However, in a real life environment, the JIT system is subjected to various uncertain factors including stochastic processing times, variable demand and process interruption due to planned preventive maintenance. These factors seriously compromise the performance of JIT. In this paper, we present a newly developed JIT system which uses an algorithm to dynamically and systematically manipulate the number of Kanbans in order to offset the blocking and starvation caused by the said factors during a production cycle. We refer to the new system as the Flexible Kanban System (FKS). We provide steps of the algorithm and demonstrate the effectiveness of FKS using a case example. For the case example, we present the solution procedure, results and discussion.
In this paper we explore the impact of sudden breakdown of the material handling system on the performance of a Traditional Kanban System (TKS). TKS, which is an element of the Just-In-Time system, is designed to operate in an ideal environment such as constant processing times and uninterrupted processing. However, in a real life environment, the TKS could be subjected to various unpredictable factors including stochastic processing times and process interruption due to equipment failure. These factors would seriously strain the performance of TKS. We consider a TKS in which some stations are dependent on a material handling system to move parts between them. We study the effect of a sudden breakdown of such a material handling system on the performance of the TKS. In addition, we also study a newly developed Kanban system (which dynamically and systematically manipulates the number of Kanbans in order to offset the blocking and starvation caused by these factors during a production cycle) under the same conditions. We refer to the new system as the Flexible Kanban System (FKS). We compare the overall performances of the TKS and FKS by considering a variety of cases. We present the solution procedure, results and discussion for these cases.
This paper proposes a methodology to improve the process time required for the component placement process of the printed circuit board assembly with an X-Y positioning table. The methodology is applied to a previously published subproblem as well as a real-life working board configuration. Even with high speed assembly machines placing in excess of 40,000 components per hour (cph), process improvements are possible. Concentrating on the path planning portion is a valid method to increase efficiency, but as the problem approaches optimality, the planning process essentially becomes counterproductive due to the time and effort required. This paper presents an algorithm for system time improvement for use after a board placement path is established. In order to take advantage of the machine's capability to move the positioning table along each axis simultaneously, the board is physically rotated with respect to the positioning table. This paper steps through the theoretical development of a method to determine the proper angle of rotation, first for a regular "rectangular" board configuration, and then for a more applicable "general" configuration.
In recent years there has been an appalling rate of depletion of natural resources due to an ever-increasing number of consumer goods manufactured, in turn leading to an increase in the quantity of used and outdated products discarded. From an environmental point of view, it is not only desirable to disassemble, reuse and/or recycle the materials and components from the discarded products, it can also be economically justified. This paper presents a quantitative methodology to measure the profitability for product disassembly and recycling by taking both operational and environmental factors into account. To this end, a two-stage solution model that provides a unique solution for planning component recovery from products with component commonality is presented. The objective of the component recovery model is to compute the number of products to disassemble, in order to fulfill the demand of the components, at the minimal disassembly and disposal costs. An example is presented to illustrate the methodology.
A disassembly process plan (DPP) is a sequence of disassembly tasks which begins with a product to be disassembled and terminates in a state where all the parts of interest are disconnected. Disassembly process planning is critical for minimizing the resources invested in disassembly and maximizing the level of automation of the disassembly process and the quality of the parts (or materials) recovered. In this paper, we propose an algorithm which automatically generates a disassembly PN (DPN) from a geometrically-based disassembly precedence matrix (DPM). This algorithm can be used to generate DPPs for products which contain simple AND, OR, complex AND/OR, and XOR relationships. The resulting DPN can be analyzed using the reachability tree method to generate all feasible disassembly process plans (DPPs), and cost functions can be used to determine the optimal DPP. An example is used to illustrate the procedure.
We develop a methodology for the analysis of a finite buffer tandem manufacturing system where the machines follow N-policy. We model the system using Open Queueing Networks. The throughput of the system is calculated using decomposition, isolation and expansion methodologies. The methodology is tested rigorously by using orthogonal arrays to design the experiments. t-test is used to investigate the differences between the results of the methodology and their corresponding simulation results. The comparison shows that the methodology is robust and remarkably accurate over a wide range of parameters
We consider a finite buffered queue where the queue length is controlled by shutting down and restarting the server. In particular, we analyze the problem with the inclusion of holding costs for customers (or items) whereas previous research concentrated on the case without holding costs. To study the effect of holding costs, we first establish some stochastic comparisons that permit us to compare different operating policies. In addition to these structural results, we also present new results on the queue length distribution for the system. Our methods enable us to obtain the queue length distribution in closed form for phase type service distributions. As a consequence, we provide extensive numerical examples over a range of the problem parameters which uncover some intriguing properties of the optimal revenue depending on the service time distribution.
With recent awareness and emphasis on quality, system reliability and maintainability have been getting a lot of attention. Increasing competition in the marketplace as well as general dependence on highly complex systems further highlight the importance for the need of reliability and maintainability. Moreover, prior knowledge of reliability and maintainability allows reasonably dependable predictions to be made of things like marketability, parts requirements, warranty costs etc. This paper presents an introduction to reliability and maintainability.
We develop a methodology for the analysis of finite buffer manufacturing systems with unreliable machines and arbitrary topology. We model the system using Open Queueing Networks. Decomposition, isolation and expansion methodologies are used to calculate the throughput of the system. The methodology is tested rigorously. Orthogonal arrays are used to design the experiments in order to cover a large experimental region. The results of these experiments are compared to their corresponding simulation results. In order to investgate the differences between the simulation results and the results of the methodology, t-tests are carried out. When tested over a wide range of parameters, the results show that the methodology is remarkably accurate and robust.
Electroplating is a major process in the manufacturing of printed circuit boards. Scheduling the movement of material handling hoists for electroplating processes is generally known as the Hoist Scheduling Problem (HSP) and has been proven to be NP-complete. The objective of HSP is to find a cyclic sequence of hoist moves that maximizes the production throughput. For the past two decades, various optimization and heuristic techniques have been proposed to solve the problem. However, these methods are often limited to the elementary problems. Recently, artificial intelligence (AI) approach using constraint logic programming has been applied to solve the cyclic HSP but did not consider problems with duplicated process tanks. In this paper, we apply constraint satisfaction to solve HSP with duplicate process tanks. A binary search procedure is proposed and a tighter bound to the cycle length is introduced to reduce the computation effort. The proposed algorithm can be easily implemented on any personal computer with reasonable performance so as to be useful on the shop floor. Finally we present results for several benchmark examples.
In recent years there has been an appalling rate of depletion of natural resources due to an ever-increasing number of consumer goods manufactured, in turn leading to an increase in the quantity of used and outdated products discarded. From an environmental point of view, it is not only desirable to disassemble, reuse and/or recycle the components and materials from the discarded products, in many cases it can also be economically justified. This paper presents a quantitative methodology for product disassembly and recycling by taking both operational and environmental factors into account. To this end, a mathematical programming model that provides a unique solution for planning component recovery from products with component commonality is presented. The objective of the component recovery model is to compute the number of products to disassemble, in order to fulfill the demand of the components, at the minimal disassembly and disposal costs. A case study is presented to illustrate the methodology.
We consider a manufacturing system with finite buffer and arbitrary topology where a machine takes a vacation (i.e. is unavailable for processing due to the processing of secondary jobs or maintenance of machines) of random duration every time the machine becomes idle. To this end, we develop an approximation (analytical) methodology to calculate the throughput of the system using queueing networks together with decomposition, isolation and expansion methodologies. The methodology was tested rigorously covering a large experimental region. We used orthogonal arrays to design the experiments in order to keep the number of experiments manageable. The results obtained using the approximation methodology were compared to simulation results. The t-tests carried out to investigate the differences between the two results showed that the proposed methodology is very accurate as well as robust.
This paper addresses the importance of planning for disassembly as a response to recent regulatory trend and increased consumers' environmental awareness. For the disassembly of a product, the size of the problem dramatically increases as the number of components increases. When planning involves multiple products, yet another layer of complexity is added to the problem. The proposed technique, case-based reasoning (CBR), is an approach to solve the problem of multiple products disassembly. CBR is a technique that allows a process planner to establish, store, reuse, and/or adapt the solution to past disassembly problems. Once a product has been initialized (established) and maintained in the case memory, a planner is allowed to search and retrieve (reuse) the product's disassembly process plan. If a product does not exist in the case memory, the planner can retrieve a product structure that is similar to the new one and make the necessary adjustments (adaptation). A derived plan can then be quickly optimized for processing. This way, a disassembly facility can operate efficiently in a multiple products reclamation system.
Developing models to study remanufacturing production processes is a difficult task due to the complexity of remanufacturing systems. A typical remanufacturing system consists of disassembly, remanufacturing, and reassembly operations. Previous models have been limited to simulation studies that require extensive time to develop, execute, and analyze. Given the complexity of most simulation models, it may be difficult to isolate specific cause-and-effect relationships required to develop effective manufacturing control techniques. We develop a queuing network model for a remanufacturing production system and present approximate solutions. A decomposition approach is used where the remanufacturing production process is decomposed into: a disassembly segment, a remanufacturing operations segment, and a reassembly segment. The disassembly segment is modeled as a simple queue, the remanufacturing segment is modeled as an open Jackson network, and the reassembly segment as a kitting process. The queuing model is compared with a simulation network model and error bounds are set. The advantages of an analytic model are discussed and applications are presented.
The complexity of planning for disassembly, as well as the time required, increases with the number of components in a product. Furthermore, in dealing with a multiple products situation, it is important to have the capability to create disassembly process plans quickly in order to prevent interruptions in processing. The application of case-based reasoning (CBR) approach in planning for disassembly can go a long way in avoiding interruptions in processing. CBR is a technique that allows a process planner to rapidly retrieve, reuse, revise, and retain the solution to past disassembly problems. Once a planning problem has been solved and stored in the case memory, a planner can retrieve and reuse the product's disassembly process plan any time in the future. The planner can also adapt an original plan for a new product that does not have an existing plan in case memory. Following adaptation and application, the successful plan is retained in the case memory for future use. In this paper, an approach to solve the problem of multi-product/multi-manufacturer disassembly is presented. The focus is on the procedures to initialize a case memory for different product platforms, and to operate a CBR system which can be used to plan disassembly processes.
In this paper, we discuss a new problem, the disassembly line balancing problem (DLBP), which can simply be defined as the optimum assignment of disassembly tasks to the disassembly workstations under the condition that the precedence relationships among the tasks are not violated. The objectives are to meet the demand and to utilize the disassembly line as efficiently as possible. We present a systematic approach to solve a simple DLBP. An example is also presented to illustrate the approach.
In this paper, we introduce an open queueing network with finite buffers to model a remanufacturing system. The system consists of three modules, viz., a testing module for returned products, a disposition module for non-reusable returns and a remanufacturing module. We analyze the network using the decomposition principle and the expansion methodology. The model has been shown to be very rigorous and remarkably accurate. An example is presented to illustrate the use of the model.
This paper presents a technique to analyze the design efficiency of a product at both ends of the life-cycle. The design efficiency is measured using a Design for Disassembly Index (DfDI). DfDI uses a disassembly tree (DT) which relies on product's structural blueprint. The DT can be used to identify precedent relationships that define the structural constraints in terms of the order in which components can be retrieved. DfDI can be used to compare the merits and drawbacks of different product designs. The index offers designers with an important measure to help improve future products. We provide a comprehensive procedure for developing the index and demonstrate its application through an example.
In this paper, we develop an open queueing network model to obtain the total cost performance of a remanufacturing system. The model is developed using decomposition, isolation and expansion methodologies. The remanufacturing system consists of three modules, viz., a disassembly and testing module for returned products, a disposition module for non-reusable returns and a remanufacturing module. The model is thoroughly tested using an experimental design based on orthogonal arrays. The results show that the model is very robust and remarkably accurate.
In this paper, we propose a methodology to determine the number of products to disassemble in order to fulfill the demand of various components for remanufacturing. A modification of the material requirements planning (MRP) technique is used to determine the number of components needed to remanufacture products in different time periods. An example is presented to illustrate the technique.
We present a generalized stochastic Petri net (GSPN) model of a single-process, single-product, manufacturing facility with a supply management system for multiple vendors controlled by kanbans. Its properties are investigated and the effect of the vendors' activities on the system are discussed.
Disassembly is a systematic method for separating a product into its constituent parts, materials, or other groupings. While disassembly process planning is critical in minimizing the amount of resources (e.g., time and money) invested in the disassembly process, solving the disassembly line balancing problem (DLBP) is crucial to maximizing the utilization of the line on which the products are taken apart. We propose a systematic approach for solving the DLBP by considering several important factors unique to disassembly when balancing the disassembly line. An example is given to illustrate the approach.
This paper presents a technique to analyze the efficiency of designing electronic products for the environment. The efficiency of each design is indicated using a Design for Disassembly Index (DfDI). DfDI uses a disassembly tree (DT) which relies on the product's bill of materials as its structural blueprint. DfDI can be used to compare the efficiency among alternative designs, identifying the best alternative for a product retirement plan. In addition, the index offers designers with an important measure to help improve future products.
Just-In-Time (JIT) systems were originally designed for deterministic production environments such as constant processing times and smooth and stable demand. However, once implemented, JIT is fraught with numerous types of uncertainties, including variations in processing time and demand, planned interruptions such as preventive maintenance and unplanned interruptions such as equipment failure. These uncertainties lead to lowered production throughput, decreased machine utilization, increased order completion time and greater backlogs and overtime requirements. In this paper, we introduce a newly developed system, which we refer to as the Flexible Kanban System (FKS), to cope with uncertainties and planned/ unplanned interruptions. We demonstrate the superiority of the new system by considering four case examples covering various uncertainties, conducting numerous studies and comparing the overall performances of the FKS with that of the traditional JIT system. In all the cases considered, the performance of the FKS was, indeed, superior to that of the traditional JIT system.
Modeling and analysis of JIT under realistic assumptions presents a number of challenges, including the ability to model blocking and starvation, the ability to conduct both qualitative and quantitative analysis, and the ability to model control policies. Petri nets (PNs) have recently emerged as a promising approach for modeling manufacturing systems. PNs are a graphical and mathematical technique useful for modeling concurrent, asynchronous, distributed, parallel, nondeterministic, and stochastic systems. PN models can be analyzed to determine both their qualitative and quantitative properties. In this paper, we use stochastic, colored PNs (SCPNs) to model JIT system and analyze the impact on system performance of two different kanban control policies: a traditional kanban system (TKS) and a recently introduced flexible kanban system (FKS) policy. In TKS, the number of kanbans is fixed throughout the production cycle; the control problem is to determine the optimal number of kanbans. In FKS, the number of kanbans is systematically manipulated during the production cycle to improve system performance; the control problem is to determine the minimal number of kanbans and when and by how much to manipulate the kanbans. The resulting models are shown to be live and bounded, and their performance compared for a variety of operating scenarios. The model can be extended to build models of arbitrary size.
In this paper we consider arbitrary topology open manufacturing (queueing) systems with finite buffers and N-policy. N-policy involves a queueing system in which the machine (server) is assigned to alternative jobs when it becomes idle and becomes available only after the queue builds up to a predetermined level of N jobs. We use the decomposition, isolation and expansion methodologies to calculate the throughput of the system. The methodology is tested rigorously by using orthogonal arrays to design the experiments in order to cover a large experimental region. The results of the methodology are compared to simulation results. To this end, we also develop a simulation model (which in itself is quite challenging). The differences in the two results are investigated using t-tests. Based on the results, the methodology proves to be remarkably accurate and robust over a broad range of parameters.
In this paper, we present a technique for analyzing the design efficiency of electronic products, in order to study the effect of end of life (EOL) disassembly and disposal on the environment. The design efficiency is measured using a Design for Disassembly Index (DfDI). DfDI uses a disassembly tree (DT) which relies on the product's structure. The DT can be used to identify the precedent relationships that define the hierarchy of the product's structure (which in turn, represents the order in which components can be retrieved). DfDI can be used to analyze the merits and drawbacks of different product designs. The index offers designers with an important measure to help improve the future products. We provide a comprehensive procedure for developing the index and demonstrate its application through an example.
Product manufacturers are subject to numerous and often contradicting demands. The array of demands include functionality, cost effectiveness, appeal, durability, timeliness and maintainability. Increasingly significant is the demand for manufacturers to keep up with consumer's appetite for cutting-edge technology. In order to be competitive, manufacturers have to develop, produce and deliver the latest production models to consumers at an astronomical rate. In turn, consumers are replacing outdated products even though they are still operational. Products ranging from electronics, home appliances, to automobiles are the growing list of used products scrapped that requires proper disposal (i.e., incineration and landfill). Furthermore, since the incinerators and landfills are becoming limited and, thus, costly to use, there is an urgency to identify alternative means of products' disposal. With increasing significance, environmental organizations and legislatures are calling on manufacturers to take responsibility, in part, for the disposal of products in an environmentally benign manner. In many parts of the world, manufacturers are urged to take-back their products and recover the components and materials at the end of their useful lives. With profits being the most important motivation for businesses, manufacturers have to face the challenge in recovering components and materials that have some residual value, while bearing the operational costs incurred in the recovery operations. The focus these days is on developing new operational techniques on planning for component and material recovery-to reduce, reuse and recycle products and their components in the most cost efficient manner. Considerations such as the potential for component or entire assembly reclamation, reuse, remanufacture, or recycle are important at this stage. Even energy recovery to some extent can be planned if the product can be taken apart for safe incineration, therefore allowing energy to be extracted from the remaining waste. One particular requirement of a component and material recovery system is the need for disassembly prior to their retrieval, reuse, recycling, and disposal. Disassembly is the process of systematic separation of the constituent components from a product assembly. Disassembly provides the opportunity to recover and isolate valuable components and/or materials from scrap by means of manual or automatic separation processes. With stringent take-back regulations and the corresponding economic consequences, no longer is disassembly an alternative to traditional disposal methods like incineration and landfill-it is a necessity. The disassembly operations have become increasingly complex since we have to consider the return flow of products, the outgoing supply of disassembled components, and the disassembly process itself. The operational aspect of disassembly has become a problem of such importance that it has accumulated a rich body of knowledge for the study of planning for disassembly processes. Questions that are central to the disassembly operations include: How to most efficiently collect the products? How to disassemble? What is the optimal amount to disassemble? and How to dispose of unused disassembled components?
The desire to achieve manufacturing excellence has become the driving force behind the development and implementation of various innovative techniques. Foremost amongst the more successful ones is the Just-In-Time (JIT) technique. The JIT philosophy evolved from a number of principles such as the elimination of waste, reduction of production cost, total quality control and recognition of employees' abilities. The objective of JIT is the production of defect free goods in the required amount at the right time. JIT was first used in the Toyota production system. Since that time, a large number of manufacturing companies in the U. S. have either implemented the JIT system or are seriously considering implementing it. A survey conducted by the American Production and Inventory Control Society (APICS) showed that, as early as the early eighties, 250 companies out of the Fortune 1000 in North America had considered using some form of JIT in their manufacturing processes or services. Since then, the popularity of JIT has spread to medium and even small companies. This trend is caused by the success of the Japanese in the area of manufacturing. The Kanban system is an element of Just-In-Time system that has captured the most attention of researchers. Kanban is a Japanese word that means "visible sign" or card. An advantage of the Kanban system is its ability to control production. Other advantages include its simplicity in production scheduling, reduced burden on operators, ease of identification of parts by the Kanbans attached to the containers and substantial reduction in paper work. The Kanban system is viewed as an information system. The Kanban contains information such as the Kanban type, component name and number, the station location and the destination station. There are different types of Kanbans for different functions. These include Withdrawal Kanbans, Production Kanbans, Supplier Kanbans, Signal Kanbans, Common Kanbans, Tunnel Kanbans, Express Kanbans and Emergency Kanbans. Since, the JIT system was designed for a deterministic environment (e.g., constant processing times and smooth and stable demand), its performance is optimum in that environment. However, once implemented, JIT is fraught with numerous types of uncertainties such as processing time and demand variations, breakdowns and other types of planned or unplanned interruptions. Even though, in most of the applications, the number of Kanbans are usually held fixed, there are times when supervisors, on an ad hoc basis, increase or decrease the number of Kanbans depending on whether the system is experiencing shortages or inventory build-up. In general, they do not manipulate the number of Kanbans systematically. However, in a stochastic and uncertain environment, it is beneficial to fluctuate the number of Kanbans systematically during the production cycle to compensate for the discrepancies introduced by the unpredictability. A new system which allows systematic fluctuations in the number of Kanbans, known as the Flexible Kanban System (FKS), does just that.
Electric vehicles (EVs) are being driven throughout the U.S. When these vehicles reach the end-of-life (EOL), their use of lighter materials such as aluminum and composites and the presence of a battery pack may impact vehicle recycling. Using goal programming techniques and a model of the automobile recycling infrastructure, materials streams and process profitabilities are tracked for General Motors' EV, the EV1. The significant amount of aluminum found in the EV1 will make it very profitable to shred as long as waste disposal costs remain low. However, there is considerable uncertainty regarding the profits that the disassembler will achieve. Although under typical market conditions, the lead acid battery pack will generate significant revenue for the disassembler and battery recycler, the immature and possibly limited market for used EV parts may result in reduced profits for the disassembler. Even so, the results indicate that the current automobile recycling infrastructure in the U.S. will be viable for recycling EVs.
In this paper, a generalized queueing system with finite source, N-policy with startup time and warm spares is considered. A closed form stationary distribution of the number of customers in the system for such a system is not attainable. However, it is possible to derive closed form expressions for recursively calculating the stationary distribution. With some modification, this model can accommodate server vacations with exhaustive service discipline. To this end a generalized model to accommodate the cases of multiple vacations, single vacation and hybrid multiple/single vacation schemes is considered. Closed form expressions for the vacation models are achievable and are presented. An efficient algorithm is presented to find the stationary probability distribution of the number of customers in the system as well as the performance measures of these models. The algorithm is an extremely powerful generalized methodology which can solve dozens of standard and non-standard queueing problems. Several special cases are considered. Some previously published results are shown to be special cases of the more general results derived here. An example and some numerical results are also presented.
We present a generalized stochastic Petri net (GSPN) model of a single-process, single-product, manufacturing facility with a supply management system for multiple vendors controlled by supplier kanbans. We consider a scenario with two vendors and obtain numerical results using the GSPN model. Its properties are investigated and the effect of the vendors' activities on the system are discussed.
We develop an analytical methodology for the analysis of a tandem manufacturing flow line with finite buffers and unreliable machines. The flow line is modeled using Open Queueing Networks. The methodology uses decomposition, isolation and expansion methodologies to calculate the throughput of the flow line. The methodology is tested rigorously. In order to cover a large experimental region, orthogonal arrays are used to design the experiments. The results of these experiments are compared to their corresponding simulation results. t-test is carried out to investigate the differences between the simulation results and the results of the methodology. The results show that the methodology is robust and remarkably accurate over a wide range of parameters.
Environmentally Conscious Manufacturing and Product Recovery (ECMPRO) has become an obligation to the environment and to the society itself, enforced primarily by governmental regulations and customer perspective on environmental issues. This is mainly driven by the escalating deterioration of the environment, e.g., diminishing raw material resources, overflowing waste sites, and increasing levels of pollution. ECMPRO involves integrating environmental thinking into new product development including design, material selection, manufacturing processes and delivery of the product to the consumers, plus the end-of-life management of the product after its useful life. ECMPRO related issues have found a large following in industry and academia who aim to find solutions to the problems that arise in this newly emerged research area. Problems are widespread including the ones related to life cycle of products, disassembly, material recovery, remanufacturing, and pollution prevention. In this paper, we present the development of research in ECMPRO and provide a state-of-the-art survey of published work.
New electronic products are usually compact and equipped with the latest technology. They are replacing outdated ones at an astronomical rate. Ironically, a large number of outdated products are often in excellent condition. Rapid product development coupled with consumer appetite for latest models of products, have caused consumers to discard outdated products even though they are still operational. This in turn leads to an increase in the quality of used and outdated products scrapped. Products made with reusable components, retrieved from discarded electronic products, are sometimes not only cheaper but also better. For example, electronic chips recovered from outdated computers could prove to be more reliable than the new chips (if care is taken to protect them from thermal damage during removal), because the reused chips would have survived the "burn-in" period. In addition, because retrieved parts are often classified as scrap, manufacturers could obtain them at a below-market cost. Reuse and recycling of electronic products have not only been driven by the return on capital concept, but also by the return to nature concept. Environmental awareness and recycling regulations have been putting pressure on manufacturers and consumers, forcing them to produce and dispose of products in an environmentally friendly manner. In many parts of the world, and especially in Europe, the regulations are becoming more stringent and manufacturers are required to recycle their products at the end of their useful lives. Current trends hold the promise for companies to do well economically as well as being environmentally friendly while meeting the impending regulations. To benefit from this new found environmentalism, electronic manufacturers have to explore possible alternatives for designing environmentally benign products. This article provides an overview of the state-of-the-art techniques for Design for Disassembly, Reuse and Recycling.
In this paper, the main focus is on an optimization approach to determine the supply of a variety of products to recover (at the end of their lives) in order to fulfill the demand of an assortment of components, and have an environmentally benign policy of minimizing waste generation. When the problem is solved, it gives the number of each product type to be disassembled in order to fulfill the demand of components needed at minimal cost. From the supply chain perspective, this would result in minimal inventory requirements at both ends of the reverse logistics chain, viz., at the end-of-life (EOL) products end and at the disassembled components end.
We present a new Kanban system specifically developed for material control and scheduling in a disassembly environment. We briefly highlight the differences between the new (modified) and the traditional Kanban system. We assert that in the disassembly environment, the Kanban system is superior to the "push" system currently practiced in industry. To that end, we consider a case example and test its performance by experimenting with several different scenarios. In all instances, the Kanban system outperforms the "push" system.
A remanufacturing system is prone to inefficiencies because of built-in uncertainties and complexities of the nature of the operations. One way to improve the performance of the system is to distribute a given number of available buffer slots among the stations in a strategic manner. In this paper we present a near optimal buffer allocation plan (NOBAP) specifically developed for remanufacturing systems. To this end, we introduce an algorithm that analyzes the system using an open queueing network with finite buffers and unreliable machines. In order to analyze the queueing network, we use the decomposition principle and expansion methodology. The results obtained by using the algorithm are compared with the ones found using the exhaustive search. The results show that the NOBAP is very rigorous and remarkably accurate.
In this paper, we present a two-stage kanban control model to study a two-echelon hybrid inventory system with disposals. We then express the expected cost function for the inventory system in terms of the performance measures obtained from the analysis of the kanban model. Finally, we present the results of a series of experiments that are conducted to measure the impact of the kanban sizes on the expected total cost of the system.
In this paper, we introduce a generalized stochastic Petri net (GSPN) to model a single-process, single-product, Just-In-Time (JIT) manufacturing facility with a supplier kanban system for multiple vendors. Its properties are investigated and the effects of the vendors' activities on the system are discussed using ANOVA.
In this paper, we present a near optimal buffer allocation plan (NOBAP) specifically developed for cellular remanufacturing system. To this end, we propose an algorithm for the analysis of the system. The algorithm analyzes the system using an open queueing network with finite buffers and unreliable machines. The remanufacturing system considered here consists of three stations, viz., a disassembly and testing station for returned products, a disposition station for non-reusable returns and a remanufacturing station. In order to analyze the queueing network, we use the decomposition principle and expansion methodology. Finally, the buffer allocation algorithm distributes the given number of available buffer slots among the three stations to optimize the system's performance.
In this paper, alternative price policies for a hybrid system with disposals are studied. An expected revenue model is constructed with different revenues for remanufactured and new items. The model is then used to study the effectiveness of the price policies under different scenarios. To this end the value iteration method is used in calculation of the average total revenues for their respective scenarios.
This paper presents a modified Kanban (pull) system for the disassembly environment and compares it to its "push" counterpart. Various scenarios are explored using a case example. For the scenarios, the assumptions, input data and results are presented. The study clearly demonstrates the effectiveness of the modified Kanban system over the push system.
This paper focuses on Supply Chain Optimization system for reverse logistics of products that are taken back for disassembly and retrieval of reusable components for remanufacturing. The goal is to provide a cost efficient way in which manufacturers can reclaim various models of a product for remanufacturing.
This paper focuses on Supply Chain Optimization system for reverse logistics. The solution approach employs an adaptation of the Materials Requirements Planning (MRP) technique, termed as Components Requirements Planning, to determine the number of components needed to remanufacture products in each time period throughout the planning horizon.
The last few years have seen a tremendous growth in the demand for durable consumer goods. The rapid development and improvement of products have given rise to additional demand resulting in shortened lifetime of most products. This in turn has increased the quality of used products scrapped. The bulk of the scrap comes from automobiles, household appliances, consumers electronic goods, and at an increasing rate from computers. There is an urgency to find alternative ways to dispose of products because of the alarming rate at which the landfills are being used up. In addition, environmental awareness and recycling regulations have been putting pressure on many manufacturers and consumers, forcing them to produce and dispose of products in an environmentally responsible manner. In many parts of the world (and especially in Europe), the regulations are becoming more stringent and manufacturers are urged to take-back and recycle their products at the end of their useful lives. Furthermore, sometimes they are also urged to use recycled materials whenever possible. This article presents an introduction to environmental issues in manufacturing systems.
Disassembly is the process of systematic removal of desirable constituent parts from an assembly for reuse or recycling. In addition, disassembly provides the opportunity to recover valuable materials or isolate poisonous substances from scrap by means of manual or automatic separation process. It is an alternative to traditional disposal method like incineration and landfill. Besides it adds value to the existing product because, instead of throwing it away, the products can be disassembled for refurbishment, reuse, or recycling.
There are essentially two types of disassembly, viz., non-destructive and destructive. Non-destructive disassembly removes desirable parts from an assembly while ensuring that there is no impairment to the parts. Destructive disassembly, on the other hand, is the process of separating similar metals and materials from an assembly in order to sort each material type for recycling.
Dry sorting is a method of separating a mixture of different materials by distinguishing the differences in material properties, such as size or density. Materials with different particle sizes can be sorted by screening the smaller ones from the larger ones.
The term 'green' is commonly used to denote an environmental friendly activity. The objective of green manufacturing is to design and produce products in an environmentally benign manner. The characteristics of an environmentally benign product ranges from the use of recyclable materials or recycled materials wherever possible, the use of fixtures and/or components that can be easily taken apart, or the design of product structure to allow ease of disassembly.
According to the Society of Environmental Toxicology and Chemistry, Life Cycle Assessment is defined as "An objective process to evaluate the environmental burdens associated with a product, process or activity by identifying and quantifying energy and materials used and wastes released to the environment, to assess the impact of those energy and materials uses and releases on the environment, and to evaluate and implement opportunities to affect environmental improvements. The assessment includes the entire life cycle of the product, process or activity, encompassing extraction and processing of raw materials, manufacturing, transportation and distribution, use/reuse/maintenance, recycling and final disposal."
Remanufacturing transforms durable products that are worn, defective, or obsolete into functional products through disassembly, cleaning, refurbishment and replacement of components, reassembling, and testing. The range of services can be: repair or replacement of broken or unreliable parts, system upgrade, minor system modification, routine system inspection, and product refurbishment to enhance appearance. A remanufactured product may retain its old form, or may lose the original identity all together.
Petri nets (PNs) have recently emerged as a promising approach for modeling flexible and automated manufacturing systems. PNs are a graphical and mathematical modeling technique that is useful for modeling concurrent, asynchronous, distributed, parallel, nondeterministic, and stochastic systems, as attested by a steady stream of papers which appeared throughout the 1980s. Since 1989, there has been an explosion of interest in using PNs to model, simulate, and analyze manufacturing systems. This article presents an overview of PN theory (including classical PNs, timed PNs, stochastic PNs, and high-level or colored PNs) and PN modeling of manufacturing systems (including flow lines, automatic transfer lines, job shops, flexible manufacturing systems, and assembly systems).
Asynchronous systems are systems in which sequences of events or activities occur in parallel (concurrently) and where there is no direct synchronization of the sequence of events. Synchronization may occur at certain points in the process (for example, the final assembly), but the activities that occur in parallel may not be synchronized. Job shops and flexible manufacturing systems frequently contain asynchronous processes. These asynchronous processes may occur when more than one type of product is being manufactured in different parts of the production facility or when an order for a product generates multiple orders which go to different parts of the production facility.
In contrast to asynchronous systems, in synchronous systems (such as automatic transfer lines) each stage in the production line is synchronized with the other stages in the line, and parts move from one stage to the next, at the same time.
Blocking in a manufacturing system occurs when a machine cannot unload a part because its output buffer is full or there is no material handling equipment available to move the part to the next work center. Blocking can also occur for material handling equipment (e.g., automated guided vehicles (AGV) and robots) when they cannot move forward because another piece of equipment is in the way or the next operation is not ready. Blocking may either be temporary or permanent. In the former case, a machine is blocked until space opens up in the output buffer, or an AGV is blocked until the obstruction is gone. In the latter case, the machine or AGV may be deadlocked and require manual intervention.
The events and activities that occur in a manufacturing system may take place sequentially (one at a time) or concurrently (in parallel). In sequential systems, the events or activities take place one after the other. For example, assembling a computer may require the following sequence of steps: position the case; position the motherboard; secure the motherboard to the case with screws; insert and secure the specified components (hard drive, CD-ROM, floppy drive); attach the cables from the components to the motherboard; insert additional cards and attach their cables; put on and secure the cover. In contrast, in concurrent systems, events and activities take place in parallel. For example, the components that are used in the assembly of the computer in the previous example, may themselves be prepared for assembly in different parts of the plant, which operate concurrently.
Obviously, most systems are not entirely sequential or concurrent; rather, they are mixed, with some portions operating sequentially and some operating concurrently. At the most basic level, even when production of an individual product is a purely sequential process, there is likely to be more than one item undergoing production at the same time; i.e., the items are being produced concurrently. In more complex systems, an individual item may be produced in a mixed mode, where some of the production steps occur sequentially and others occur concurrently.
In a manufacturing system, a deadlock occurs when more than one job requests the same resource. For example, consider the case of a machine that has a single input/output buffer with a capacity of one. Assume that an incoming job is placed in the machine's buffer, and after the machine removes the job to work on, another job is placed in the buffer. When the machine completes work on the first job, it has nowhere to put the completed part, since the buffer is full. In this case, the machine, the worked part, and the incoming part are all deadlocked. In many real systems, the only way to recover from this situation is to manually intervene to release the shared resource (e.g., a buffer). In some cases, this may necessitate shutting down a portion or all of the system.
This term is used with mathematical or simulation models of systems. A discrete-event system is one in which events occur at discrete points in time, there exist a discrete set of states, and the states change at discrete points in time. For example, an automobile manufacturing facility is a discrete-event system. The following events occur at discrete points in time: orders arrive, a production step begins, a production step ends, a machine breaks down, a machine comes back on line after repair, and so on. Similarly, at any point in time, there are a discrete number of orders pending, machines operating, and automobiles in production.
Examples of discrete-event systems in manufacturing include production lines, flow and assembly lines, job shops, flexible manufacturing systems, automated manufacturing systems, and so on. Other examples of discrete-event systems include any queueing type of system (e.g., a bank, grocery store, or tollbooth), computer systems, communications systems, and traffic systems.
In contrast in continuous systems, the system variables change continuously with respect to time. Examples of continuous systems in the manufacturing domain include chemical manufacturing (e.g., batch processing and mixing), annealing processes, and other processes which involve heating or cooling, and chemical reactions.
Makespan is the total time required to process all the jobs in an order or in a batch. Makespan is typically measured beginning from the time the first job in an order enters the manufacturing process through the time the last job in the order is completed.
A correct system design is one in which the system only enters states that are desirable and does not enter states, which are undesirable. System correctness is a logical feature of the system design (i.e., it does not refer to system performance issues such as throughput, makespan, and so on). In manufacturing systems, we wish to design systems that are deadlock free, do not have buffer overflows, and do not violate any precedence relationships (e.g., the production steps).
This article describes the mechanism of a Traditional Kanban System (TKS). TKS, which is an element of the Just-In-Time system, is designed to operate in an ideal environment such as constant processing times and uninterrupted processing. However, in a real life environment, the TKS could be subjected to various unpredictable factors such as stochastic processing times, variable demand and process interruption due to equipment failure. These factors would seriously strain the performance of TKS. In TKS, the number of Kanbans are always held fixed during the production cycle. Although, it is well known that supervisors, from time to time, on an ad hoc basis, increase or decrease the number of Kanbans depending on whether the system is experiencing shortages or inventory build-up, they do not use any specific technique to accomplish this. Flexible Kanban System (FKS) is a newly developed Kanban system which dynamically and systematically manipulates the number of Kanbans in order to offset the blocking and starvation caused by these factors during a production cycle. This article describes the mechanism of FKS and gives examples of its application.
Adaptive kanban control is a dynamic kanban control (DKC) policy, the goal of which is to systematically add and remove kanbans so as to achieve improved system performance. Under a DKC policy, kanbans are added (released) when needed to improve system performance and removed (captured) when they are no longer needed or when their presence will result in lowered system performance. In general, we want additional kanbans when the benefits of their presence (e.g., reduced blocking and starvation, improved throughput) outweigh the costs (e.g., increased WIP and operating costs). In the adaptive kanban control policy, the release and capture thresholds are a function of demand and finished inventory levels. The thresholds are evaluated when a new demand arrives and when a part enters the finished inventory. The goal of this DKC policy is to reduce holding and inventory cost in the presence of extreme variations in demand and lead times.
Kanban is the Japanese word for "visible sign" or card. Kanbans are used to control the flow of materials, parts and product in just-in-time (JIT) or pull-production systems. During the 1990s, researchers developed a class of kanban control policies, known as dynamic kanban control policies, which systematically manipulate the number of kanbans in the production system. Some of these policies use the concept of maintaining a minimal or "base" number of kanbans in the system. They may be augmented by additional kanbans as warranted; the extra kanbans may be removed according to the dynamic control policy. The base kanbans are always present in the system and represent the minimum number of kanbans in the system.
Many system control policies attempt to generate plans to cover some future period of time. This period of time is known as the decision horizon for the policy; it is the period of time for which the plan is active. Typically, the policy will be reevaluated prior to the end of the decision horizon, as new data become available; in this case, a new plan is implemented for the specified decision horizon. For example, some dynamic kanban control (DKC) policies plan the release and capture of kanbans over a specified period of time.
A deterministic production environment is one in which the demand, production times, maintenance times, and resource availability are predictable and have no variability. Such an environment is ideal from a planning perspective. In any event, plant supervisors strive to minimize the internal sources of variation (e.g., machine breakdown, maintenance schedules, staff turnover, etc.), even when they cannot completely control the sources of external variation (e.g., demand).
Dynamic kanban control (DKC) is the general name for a class of kanban control policies whose aim is to systematically manipulate the number of kanbans in a just-in-time (JIT) system in order to achieve certain performance goals. DKC policies have been developed to cope with variations in demand, sudden equipment breakdown, and processing time variations.
Employee turnover in a company refers to the movement of workers in and out of employment. It disturbs planning and production. Employee turnover comes in several forms: voluntary separations, layoffs, discharges for disciplinary reasons, retirement, or permanent disability. Employee turnover may be defined as the ratio of the number of employees hired to replace those who have left, to the total number employees.
Flexible kanban control is a dynamic kanban control (DKC) policy. The goal of DKC is to systematically add and remove kanbans so as to achieve improved system performance. Under a DKC policy, kanbans are added when needed to improve system performance and removed when they are no longer needed or when their presence will result in lowered system performance. In general, we want additional kanbans when the benefit of their presence (e.g., reduced blocking and starvation, improved throughput) outweighs the cost (e.g., increased WIP and operating costs). The goal of these DKC policies is to minimize backlog, while holding down WIP, in the face of variations in processing time and demand, and sudden equipment breakdown.
In Kanban-based manufacturing systems, the number of Kanbans is generally held fixed during a production cycle. It is, however, well known that supervisors, from time to time, on an ad hoc basis, increase or decrease the number of Kanbans depending on whether the system is experiencing shortages (starvation) or inventory buildup (blocking). In fluctuating processing time and variable demand environments, it is beneficial to adjust the number of Kanbans during the production cycle. This type of system, which allows changes in the number of Kanbans, is termed as a Flexible Kanban System (FKS). Recent research shows how to change the number of Kanbans in order to offset blocking and starvation caused by uncertainties during a production cycle.
Kanban is the Japanese word for "visible sign" or card. Kanbans are used to control the flow of product through a production facility; they are typically used in just-in-time (JIT) or pull-production systems. Kanbans are attached to the products or to containers of products that flow through the system. A kanban may contain information identifying the part, the workstation, and the order. By keeping track of the kanbans, the system can keep track of the work in process (WIP).
In kanban-based production systems, Kanban cards in manufacturing represents a type of signal to begin production of a part. Without this signal, production cannot occur. The number of Kanban cards thus limits the amount of work-in-process inventory that can accumulate between workstations or within a cell. It can also be used to limit the amount of raw material and finished goods. Many things such as cards, containers, taped spaces, etc., can be used as kanban signals.
Production systems may be characterized as "push" production or "pull" production systems. In push production systems, raw materials and parts are pushed through the production system and the finished product is stocked to meet predicted demand. Even when production occurs in response to an order, the order triggers a release from the raw stock, which is then pushed through the production system. In pull production systems, the product is manufactured in response to a specific demand. The order is used to trigger a pulling action from the end of the production line (e.g., from the last workstation). If that workstation cannot fill the order, it requests additional units from the preceding workstation. This action continues with each following workstation requesting units from its predecessor workstation. Hence, the product is pulled through the system. When properly implemented, pull production systems result in less work in process (WIP) than do push production systems, which in turn reduces warehousing and investment costs. Pull production systems are also known as Just-in-Time (JIT) systems or Kanban systems. Pull production systems are controlled using Kanbans. The Kanbans are used to signal the pulling action from one workstation to another.
Reactive kanban control is a dynamic kanban control (DKC) policy, the goal of which is to systematically add and remove kanbans so as to achieve improved system performance. Under a DKC policy, kanbans are added when needed to improve system performance and removed when they are no longer needed or when their presence will result in lowered system performance. In general, we want additional kanbans when the benefit of their presence (e.g., reduced blocking and starvation, improved throughput) outweighs the cost (e.g., increased WIP and operating costs). The goal of the control policy is to achieve a desired order completion time and WIP.
During the 1990s, researchers developed a class of kanban control policies, known as dynamic kanban control (DKC) policies. The goal of DKC is to systematically add and remove kanbans so as to achieve improved system performance. Under a DKC policy, kanbans are added when needed to improve system performance and removed when they are no longer needed or when their presence will result in lowered system performance. In general, we want additional kanbans when the benefit of their presence (e.g., reduced blocking and starvation, improved throughput) outweighs the cost (e.g., increased WIP and operating costs). The point at which kanbans are added is referred to as the release threshold; the point at which they are removed is the capture threshold. The release threshold is the threshold in a system performance measure at which additional kanbans are released into the system. The capture threshold is the threshold in a system performance measure at which kanbans are captured and removed from the system. The specific release and capture thresholds are functions of the DKC in use.
We consider a production system with finite buffers and arbitrary topology where services are subject to interruptions in one of three ways, viz., machine breakdown, machine vacations or N-policy. We develop a unified approximation (analytical) methodology to calculate the throughput of the system. We use queueing networks together with decomposition, isolation and expansion methodologies. The methodology is rigorously tested covering a large experimental region. Orthogonal arrays are used to design the experiments in order to keep the number of experiments manageable. The results obtained using the approximation methodology are compared to simulation results. The t-tests carried out to investigate the differences between the two results show that the proposed methodology are statistically insignificant. Finally, we test the methodology by applying it to several arbitrary topology networks. The results show that the performance of the approximation methodology is consistent, robust and produces excellent results in a variety of experimental conditions.
The complexity of planning for disassembly, as well as the time required increases with the number of components in a product. Furthermore, in dealing with a multiple products situation, it is important to have the capability to create disassembly process plans quickly in order to prevent interruptions in processing. The application of case-based reasoning (CBR) approach in planning for disassembly can go a long way in avoiding interruptions in processing. CBR is a technique that allows a process planner to rapidly retrieve, reuse, revise, and retain the solution to past disassembly problems. Once a planning problem has been solved and stored in the case memory, a planner can retrieve and reuse the product's disassembly process plan any time in the future. The planner can also adapt an original plan for a new product that does not have an existing plan in case memory. Following adaptation and application, the successful plan is retained in the case memory for future use. In this paper, an approach to solve the problem of multi-product/multi-manufacturer disassembly is presented. The focus is on the procedures to initialize a case memory for different product platforms, and to operate a CBR system that can be used to plan disassembly processes.
The use of aluminum alloys in automobile production is growing as automakers strive to lower the fuel consumption of their vehicles and reduce emissions by substituting aluminum for steel in many applications. The current recycling infrastructure for end-of-life (EOL) vehicles is mature, profitable, and well suited to steel intensive vehicles; however, increased use of cast aluminum and expanded use of wrought aluminum will present new challenges and opportunities to the disassembler and shredder, who now comprise the first stages of the vehicle recycling infrastructure.
Using goal programming techniques, a model of the automobile recycling infrastructure is used to assess the materials streams and process profitabilities for several different aluminum intensive vehicle (AIV) processing scenarios. The first case simulates the processing of an AIV in the current recycling infrastructure. Various changes to the initial case demonstrate the consequences to the disassembler and shredder profitabilities when: the price of nonferrous metals changes; greater fractions of the vehicle are removed as parts; the parts removed by the disassembler have increased aluminum content; the quantity of polymer removed by the disassembler is increased; the disassembly costs increase; the disposal costs for shredder residue and hazardous materials increase; the shredder processing costs increase; and different AIV designs are considered. These profits are also compared to the profits achieved for a steel unibody vehicle to highlight the economic outcomes of introducing aluminum intensive vehicles into the existing infrastructure.
We present a pull system that uses a new approach to facilitate material control and scheduling in a disassembly environment and is called the modified Kanban system for disassembly (MKSD). We also compare the performance of MKSD to its "push" counterpart. Various scenarios are explored using a case example. For the scenarios, the assumptions, input data and results are presented. The study clearly demonstrates the effectiveness of the modified Kanban system over the push system.
This paper presents a new approach to address the problem of Planning for Disassembly (PFD). The approach is based on the Case-based reasoning technique. To assist planners to solve PFD problems, a system must have some heuristics and domain specific knowledge, which is related to the representation of the disassembly knowledge. In previous work, the authors suggested to use EMOPs (Eposodic Memory Organization Packet) for the knowledge representation of the PFD plan. This paper demonstrates the implementation of the EMOP memory model. The model has been implemented in C++, and tested. An example is presented to demonstrate the capabilities of the memory model.
Extensible markup language (XML) is a new powerful technique for Web and Internet development. It is a method of defining structured data in a text file. XML is expected to do for data what HTML has done for Web pages. XML's strength lies in its simplicity to represent data and knowledge. It can maintain hierarchical structures encountered in many systems including assemblies and disassemblies. It can also be well integrated with Java, and its Java beans. Its strength lies in its flexibility to adapt to any knowledge domain because it is a metalanguage. It is used to develop modeling languages that are tailored to specific knowledge, and specific data structures and hierarchies. This paper presents an overview of XML, followed by a proposal of an XML-based knowledge representation model for disassembly planning. An example is used to demonstrate the capabilities of the proposed XML model.
In this paper we address the operational and economical aspects of EOL computer systems at educational institutions. To this end we present an actual case study of a major university in Boston and provide an economical analysis of different options such as disposal, disassembly, recycling, reuse and re-sale of these systems. We recommend a new procedure that will improve the collection and handling processes leading to a structured decision making methodology.
In this paper, we investigate the variation in the reusable rate of cores (used products) on the performance of the remanufacturing system. The remanufacturing system considered here consists of three modules; viz., the disassembly and testing module for returned products, the disposition module for non-reusable returns, and the remanufacturing module. Each server in the system is subject to breakdown and has a finite buffer capacity. Repair times, breakdown times and service times follow exponential distributions. We model the remanufacturing system as an open queueing network and use the decomposition principle and expansion methodology to analyze it.
As the use of personal computers (PCs) increases, their short life cycle and the fact that they contain many hazardous materials means that their retirement and disposal represents a significant environmental concern. Many communities are mandating the recycling of these PCs, to recover parts and materials, and to minimize the amount of waste landfilled or incinerated. An industry to recycle these PCs is evolving to take advantage of this stream of materials. At present, PC recycling is not profitable. This paper investigates the factors that most influence the net cost to recycle PCs so that PC manufacturers, recyclers and legislators may better develop products and policies to insure that it is cost effective to recycle PCs.
The current trend of depletion of natural resources due to an ever-increasing number of consumer goods manufactured has led to an increase in the quantity of used and outdated products discarded. From an environmental point of view, it is not only desirable to disassemble, reuse, remanufacture and/or recycle the discarded products, in many cases it can also be economically justified. This situation being the motive, in recent years there have been several studies reported on disassembly, remanufacturing and/or recycling environments. Since "environmentally conscious manufacturing" is a relatively new concept that brings new costs and profits into consideration, its analysis cannot be provided by readily available techniques. This paper presents a quantitative methodology to determine the allowable tolerance limits of planned/unplanned inventory in a remanufacturing supply chain environment based on the decision-maker's unique preferences. To this end, an integer goal-programming model that provides a unique solution for the allowable inventory level is presented. The objective of the supply-chain model is to determine the number of a variety of components to be kept in the inventory while economically fulfilling the demand of a multitude of components, and yet have an environmentally benign policy of minimizing waste generation. A numerical example is presented to illustrate the methodology.
As a consequence of environmental necessities, reuse of products has recently become an important issue for production and planning. Many companies are involved in retrieving used products, where they repair, refurbish and upgrade the products in order to sell them for profit. However, the regulations for many markets do not allow manufacturers to sell remanufactured products under the same pretence as new products. Therefore, companies are forced to differentiate both the recovery and the sales activities for the remanufactured products from that of the new products. In this paper, we study the impact of this differentiation. We particularly look at the feasibility of substituting one version of the product with the other in order to satisfy the demand. In the first phase of the study, we try to find optimal switching functions for substitution decisions using a Markov decision process. In the second phase, we define several control policies and compare them with respect to the expected total cost function of the system.
Supply chain planning systems in reverse logistics present the industry with new problems that demand new approaches. The specific problem of the reverse logistics for the end-of-life (EOL) products addressed in this study is to determine the number of products to disassemble in a given time period to fulfill the demand of various components during that and subsequent time periods. We present a mathematical programming based model to solve the problem. When the problem is solved, it gives the number and timing of each product type to be disassembled in order to fulfill the demand of components needed at minimal disassembly and disposal costs. We illustrate the solution methodology with a case example.
Disassembly line is, perhaps, the most suitable way for the disassembly of large products or small products in large quantities. In this paper, we address the disassembly line balancing problem (DLBP) and the challenges that come with it. The objective of balancing the disassembly line is to utilize the disassembly line in an optimized fashion while meeting the demand for the parts retrieved from the returned products. Although, the traditional line balancing problem for assembly has been studied for a long time, so far, no one has formally talked about the DLBP. In this work, our primary objective is to address the DLBP related issues. However, we also present a heuristic to demonstrate how several important factors in disassembly can be incorporated into the solution process of a DLBP. An example is considered to illustrate the use of the heuristic.
In this paper, we present a preemptive integer goal programming approach to model the remanufacturing process to achieve various economical, physical and environmental goals. The model allows the decision-maker to determine and sort his/her goals according to their importance. A case example is presented to illustrate the use of the model.
We present a Petri net-based approach to automatically generate disassembly process plans for product recycling or remanufacturing. We define an algorithm to generate a geometrically-based disassembly precedence matrix (DPM) from a CAD drawing of the product. We then define an algorithm to automatically generate a disassembly Petri net (DPN) from the DPM; the DPN is live, bounded, and reversible. The resulting DPN can be analyzed using the reachability tree method to generate feasible disassembly process plans (DPPs), and cost functions can be used to determine the optimal DPP. Since reachability tree generation is NP-complete, we develop a heuristic to dynamically explore the ( likeliest lowest cost branches of the tree, to identify optimal or near-optimal DPPs. The cost function incorporates tool changes, changes in direction of movement, and individual part characteristics (e.g., hazardous). An example is used to illustrate the procedure. This approach can be used for products containing AND, OR, and complex AND/OR disassembly precedence relationships.
This paper presents an approach to automatically generate disassembly sequence plans (DSPs) for product recycling and remanufacturing. We first define an algorithm to generate a geometrically-based disassembly precedence matrix (DPM) from a CAD drawing of the product. The DPM is then used to generate a hierarchical disassembly tree (HDT) which represents the feasible DSPs. Generation of the HDT, i.e., generation of all feasible DSPs, is NP-complete. Thus, in order to keep the size of the HDT manageable, we control the branching and bounding processes by using two user defined variables. The first variable, w, controls the enumeration of the HDT, while the second variable, v, controls the bounding procedure in the HDT along with an evaluation function. The evaluation function incorporates tool changes, changes in direction of movement during disassembly along with individual part characteristics (e.g., high-valued parts, parts with hazardous content, etc.) The resulting HDT is called the reduced HDT (RHDT) since it only represents as many (near-) optimum DSPs as the size of v. Experimental results are presented to demonstrate the applicability and effectiveness of the methodology.
In this paper, we discuss the disassembly line balancing problem in the presence of task failures (DLBP-F). There are precedence relationships among disassembly tasks and the tasks must be completed within a given time, which is determined by the demand in a given period. However, if a task (or more than one task) cannot be performed because of some defect, some or all of the remaining tasks may be disabled due to the precedence relationships among tasks. This may result in various complications in the flow of work-pieces on the disassembly line, e.g., early-leaving, self-skipping, skipping, disappearing and revisiting work-pieces. We discuss these complications and highlight their effects on the disassembly line. The problem is to assign tasks to workstations such that the effect of the defective parts on the disassembly line is minimized. This paper presents a solution procedure to the DLBP-F. An example is provided to illustrate the approach.
This chapter presents a mathematical programming based model to solve the aggregate-planning problem for End-Of-Life (EOL) products. The goal is to provide a way in which the disassembler can reclaim various models of a product for remanufacturing in the most economical way. The approach is employed in a multi-period environment to determine the number of components needed to remanufacture products in each time period throughout the planning horizon.
When the problem is solved, it gives the number and timing of each product type to be disassembled in order to fulfill the demand of components needed at minimal disassembly and disposal costs. Hence, from the aggregate-planning perspective, this would result in minimal inventory requirements at both ends of the processing chain (viz. the supply of EOL products and the disassembled components). The application of the optimization approach is illustrated with the help of an example.
Electronic Toll Collection (ETC) holds the promise of greatly reduced congestion at toll roads, bridges and tunnels. Using automatic vehicle identification (AVI), vehicles are charged a toll as they drive non-stop through a tollbooth. Since a transition to the AVI-dominated toll plaza cannot occur overnight, a plan for the increasing use of AVI tollbooths that does not increase congestion for non-AVI vehicles is needed. Experiments using a simulation model of a typical toll plaza with varying mixes of vehicle types and tollbooth allocations provided data for four output measures: vehicle volume, average queue length, average waiting time and tollbooth utilization. Multiple linear regression analysis was then applied to fit response surfaces for each measure. The response surfaces provided the optimum tollbooth allocation for a given mix of vehicles. Two decision rules were evaluated based on their ability to specify near-optimum tollbooth allocations for varying mixes of vehicle types using vehicle volume as the output measure.
We consider a system with two discrete production lines where the output of each one can fulfill the demand for the same type of product. An example to this case is the hybrid-manufacturing environment, where a company manufactures new products and remanufactures returned products to meet the demand. The interarrival times for demand occurrences and service completions are exponentially distributed i.i.d. variables. Here, a single stage pull production control with two types of kanbans is utilized to model the system where a kanban type is dedicated for either manufacturing or remanufacturing process.
We investigate the effect of reusable rate variation of cores (used products) on the performance of buffer allocation plan for remanufacturing systems. We model the remanufacturing system as an open queueing network and use the decomposition principle and expansion methodology to analyze it.
We present a disassembly-to-order system applied in a multi-period environment where the products are taken back from the last user and/or collectors, disassembled for the retrieval of reusable items and resold (or used) in order to meet a certain level of demand for components or subassemblies. The surplus items are recycled, stored for use in subsequent periods or properly disposed of while the surplus products are only stored for subsequent periods or disposed of. We assume that the items have finite shelf lives after which they must be disposed of. Another assumption has to do with space limitation. Thus, it is not always possible to store all items in inventory even if they have positive shelf lives. We model the problem using goal programming (GP) so that there is a balance between environmental as well as economical issues. A numerical example is provided to illustrate the methodology.
Disassembly plays an important role in product recovery by allowing selective separation of desired parts and materials. The disassembly line is the best choice for automated disassembly of returned products, a feature that will be essential in the future. It is, therefore, important that the disassembly line be designed and balanced so that it works as efficiently as possible. In this paper, we address the disassembly line balancing problem and the challenges that come with it.
In this paper, we consider a system with two discrete production lines where either one can satisfy the demand for the same type of product. An example to this case is the hybrid-manufacturing environment, where a company manufactures new products and remanufactures returned products to meet the demand. The interarrival times for demand occurrences and service completions are exponentially distributed i.i.d. variables. We model this system by using a single stage pull production control with a single type of 'K' kanbans and a routing probability 'r' to distribute the detached kanban when a demand is met.
In this paper, we present a near optimal buffer allocation plan (NOBAP) specifically developed for a remanufacturing system with finite buffers and unreliable servers. In order to analyze the system we propose an algorithm that uses an open queueing network, decomposition principle and expansion methodology. The buffer allocation algorithm distributes the given number of available buffer slots among the remanufacturing system stations to optimize the system's performance.
This paper presents a multi-criteria model for the information and product flow in a disassembly-to-order environment. We assume that the used products are retrieved from the last user and/or collector and are disassembled in order to satisfy a certain demand for products, parts or materials while achieving various financial and "environmentally benign" goals.
Disassembly plays an important role in product recovery by allowing selective separation of desired parts and materials. In this paper, we address the disassembly line balancing problem (DLBP) and the challenges associated with it. We also present a methodology to solve a simple DLBP. An example is considered to illustrate the use of the methodology.
In this paper we present a new Kanban system specifically developed for the disassembly environment. We highlight its complexities and the way to overcome them. We show that the disassembly Kanban system is superior to the "push" system currently practiced in industry. An example is considered to demonstrate this superiority.
In this paper, a disassembly-to-order algorithm that incorporates the retrieval and disassembly of used products to satisfy a certain demand for products, parts and/or materials, while achieving various financial and "environmentally benign" goals, is presented. We use graph theory and goal programming to accomplish this. An example is presented to illustrate the methodology.
This paper presents a mixed integer goal-programming model that provides a solution for planning component recovery from products with component commonality. The objective of the component recovery model is to determine the aggregate number of a variety of products to disassemble in order to economically fulfill the demand of a multitude of components, and yet have an environmentally benign policy of minimizing waste generation. A numerical example is presented to illustrate the methodology.
Probability and statistics concepts have very useful applications in science and engineering. This chapter briefly introduces some of the important concepts of probability and statistics.
Product designers have always had to balance various performance requirements and cost in their designs for new products. More recently, designers have also had to consider the environmental impact of product design. Such concerns have led to the emergence of environmental design initiatives that consider the production processes and the total life cycle of the product. Examination of the recycling economics of EOL products indicates that there are many factors that have an effect on the recycling infrastructures. Using systems approaches, various tools have been developed to investigate the consequences of design alternatives and the subsequent effect on recycling economics. In this article, the results from an investigation of alternative materials in automotive bodies and the consequent recycling economics are presented as an example of the use of one of these tools.
Environmentally conscious manufacturing is an important paradigm in today's engineering practice. Disassembly is a crucial factor in implementing this paradigm. Disassembly allows the reuse and recycling of parts and products that reach their "death" after their life cycle ends. There are many questions that must be answered before a disassembly decision can be reached. The most important question is economical. The cost of disassembly versus the cost of scrapping a product is always considered. This paper develops a computational tool that allows decision-makers to calculate the disassembly cost of a product. The tool makes it simple to perform "what if" scenarios fairly quickly. The tool is Web based and has two main parts. The front-end part is a Web page and runs on the client side in a Web browser while the back-end part is a disassembly engine (servlet) that has disassembly knowledge and costing algorithms and runs on the server side. The tool is based on the client/server model that is pervasively utilized throughout the World Wide Web. An example is used to demonstrate the implementation and capabilities of the tool.
When a product reaches it's end-of-life, there are several options available for processing it including reuse, remanufacturing, recycling, and disposing (the least desirable option). In almost all cases, a certain level of disassembly may be necessary. Thus, finding an optimal (or near optimal) disassembly sequence is crucial to increasing the efficiency of the process. Disassembly operations are labor intensive, can be costly, have unique characteristics and cannot be considered as reverse of assembly operations. Since the complexity of determining the best disassembly sequence increases with the increase in the number of parts of the product, it is extremely crucial that an efficient methodology for disassembly process planning be developed. In this paper, we present a genetic algorithm for disassembly process planning. A case example is considered to demonstrate the functionality of the algorithm.
In this paper, we examine the tradeoffs between increasing the number of buffers and increasing the capacity at the remanufacturing stations under numerous circumstances on such performance measures as expected total cost, average WIP inventory, throughput and average processing (remanufacturing) time when the remanufacturing stations are operating in uncertain environments. We model the remanufacturing system using an open queueing network with finite buffers and unreliable servers. In order to analyze the queueing network, we use the decomposition principle and expansion methodology. Each server in the system is subject to breakdown and has a finite buffer capacity.
In this paper, we consider a hybrid manufacturing system with two discrete production lines. Here the output of either production line can satisfy the demand for the same type of product without any penalties. The interarrival times for demand occurrences and service completions are exponentially distributed i.i.d. variables. In order to control this type of manufacturing system we suggest a single stage pull type control mechanism with adaptive kanbans and state independent routing of the production information.
One of the first processes for preparing a product for reuse, remanufacture or recycle is disassembly. Disassembly is the process of systematic removal of desirable constituents from the original assembly so that there is no impairment to any useful component. As the number of components in a product increases, the time required for disassembly, as well as the complexity of planning for disassembly rises. Thus, it is important to have the capability to generate disassembly process plans quickly in order to prevent interruptions in processing especially when multiple products involved. Case-based reasoning (CBR) approach can provide such a capability. CBR allows a process planner to rapidly retrieve, reuse, revise, and retain solutions to past disassembly problems. Once a planning problem has been solved and stored in the case memory, a planner can retrieve and reuse the product's disassembly process plan at any time. The planner can also adapt an original plan for a new product, which does not have an existing plan in case memory. Following adaptation and application, a successful plan is retained in the case memory for future use. This paper presents the procedures to initialize a case memory for different product platforms, and to operate a CBR system, which can be used to plan disassembly processes. The procedures are illustrated using examples.
This paper outlines a framework for flexible hoist scheduling that utilizes a knowledge-based simulation system to emulate multi-hoist operations in an electroplating line and derives hoist schedule to meet production demands. The selected hoist schedule is then used by the knowledge-based controller to control the material handling hoist operations in real-time. The paper describes the overall design of the knowledge-based simulator for material handling hoist operations that forms the fundamental part of the proposed framework. Most hoist schedules in industry are experimented on-line, which are costly and time consuming. The use of knowledge-based simulation will save both cost and time. The simulator aims at improving current manual approaches so that better schedules can be identified. We demonstrate the application of the simulator to bi-directional multi-hoist line configuration that is commonly found in industry today, where loading and unloading stations are on the same end of the line, and consider other more realistic system parameters than those considered in previous studies. Three new hoist scheduling heuristics are proposed and their performances are compared with four previously studied rules. To this end, we also study the effect of several system variables such as hoist speed and inter-hoist distance on productivity. The simulation models developed are based on data from actual multi-hoist electroplating lines within several PCB manufacturing facilities.
As the use of personal computers (PCs) increases, their short life cycle and the fact that they contain many hazardous materials have a significant influence on finding economical and environmentally benign means for their retirement and disposal. Many communities are mandating the recycling of these PCs to recover parts and materials and to minimize the amount of waste landfilled or incinerated. An industry to process end-of-life PCs is evolving to take advantage of this stream of materials. At present, PC recycling is not profitable unless the recycler receives a processing fee. In this paper, the factors that most influence the profitability of end of life processing of PCs are investigated and reported so that PC manufacturers and legislators may better develop products and policies to ensure that there is a viable PCs recycling infrastructure.
One of the most responsible ways of processing end-of-life (EOL) products is to minimize their disposal to landfills by reusing, remanufacturing and/or recycling them. Often, this necessitates a certain level of disassembly. It is therefore necessary to plan disassembly processing efficiently so as to minimize costs and the amount of disposal as well as to maximize reuse, remanufacturing and recycling. In this paper, we present a disassembly-to-order system to determine the number of EOL products to process to fulfill a certain demand for products, parts and/or materials under a variety of objectives and constraints using a newly developed decision tool, called Linear Physical Programming. It addresses problems involving multiple objectives and constraints and allows the decision maker to express his/her value-system in a realistic manner for each objective of interest. The model also provides the number of items to be disassembled for remanufacturing, recycling, storage and disposal. A case example is also presented.
In this paper, we address the problem of demand driven disassembly used to determine the optimal lot-sizes of end-of-life (EOL) products to disassemble so as to fulfil the demand of various components from a mix of different product types that have a number of components and/or modules in common. We discuss two approaches, viz., (1) the disassembly graph approach that is based on the study of the disassembly of mechanical products, and (2) the component-disassembly optimisation model that focuses on parts recovery by applying the reverse bill of materials. Although elegant, the main disadvantages of these two approaches are redundancy and non-linearity respectively. To overcome these disadvantages, we propose a new method that combines the advantages of both approaches without their disadvantages. This is called the tree network model, which is a linear description of the demand driven, multiple product problem that includes commonality and multiplicity. Because of its simple structure, it can also be applied in dynamic situations, which is useful in problems that are related to production planning and inventory control in reverse logistics.
In this paper we present a multi-criteria optimization model of a disassembly to order system to determine the best combination of the number of each product type to be taken back at the end-of-life and disassembled to meet the demand for items and materials retrieved from them under a variety of physical, financial and environmental constrains so as to achieve the preemptive goals of maximum total profit, maximum sales from materials, minimum number of disposed items, minimum number stored items, minimum cost of disposal and minimum cost of preparation, in that order. When solved, the model provides the number of reused, recycled, stored and disposed items as well as the values of a host of other performance measures. A case example is presented to illustrate the model's implementation.
There are several situations in a product recovery environment where products may be disassembled for economical and regulatory reasons. The disassembly line is perhaps the most suitable setting for disassembly of large products (consisting of numerous components) as well as small products received in large quantities. In this paper, we discuss the importance of disassembly line in product recovery. The objective of the disassembly line is to utilize the available resources as efficiently as possible while meeting the demand for recovered parts. However, there are many complicating matters that need to be considered to create an efficient disassembly line. Our primary goal of this research is to discuss these issues and provide a better understanding of the complications and their effects on the disassembly line. We also demonstrate how some important factors in disassembly can be accommodated to balance a paced disassembly line by modifying the existing concept of assembly line balancing. An example is presented to illustrate the approach.
End-Of-Life (EOL) processing has recently gained enormous attention as a result of increasing environmental legislation and diminishing natural resources. Once a product reaches its EOL, there are several alternatives available for its processing, e.g., reuse, remanufacture, recycle, storage and proper disposal all requiring a certain level of disassembly. Since disassembly tends to be expensive, disassembly scheduling has been of interest as of late. However, disassembly scheduling is very complex and NP complete and therefore conventional optimization methods of tackling it are unsuitable. Heuristic approaches, on the other hand, can reduce computational time and provide reasonable results. In this paper, we propose a Multiple Objective Tabu Search methodology to solve the disassembly scheduling and process planning problem for a given EOL product. A case example is presented to illustrate the methodology.
Today, the amount of waste created by the disposal of end-of-life products has reached epidemic proportions. A major contributing factor of this phenomenon is the shortened life times of products. The most effective way to address this problem is to optimize the end-of-life (EOL) processing activities of products. There are many alternatives for the EOL processing of products, e.g., reuse, recycle, storage and proper disposal. All of these activities require a certain level of disassembly. Since disassembly tends to be a very expensive operation, special attention should be given to it for it to be efficient. For this reason, disassembly process planning, which provides a feasible sequence of disassembly, has been the focus of several recent studies. We present a disassembly-to-order system where the products are taken back from the last user and/or collectors, disassembled for the retrieval of reusable components and resold in order to meet a certain level of demand. We model it as a multi-criteria decision making problem under uncertainty using the fuzzy goal programming technique. A case example is provided to illustrate the methodology.
In recent years, the continuous growth in consumer waste has seriously threatened the environment. Realizing this, many countries have passed regulations that force manufacturers to take back their used products from consumers so that the components and materials recovered from the products may be reused and/or recycled. Disassembly plays an important role in product recovery. A disassembly line is perhaps the most suitable setting for disassembly of products in large quantities. Since a disassembly line has a tendency to generate excessive inventory, employing a kanban system can reduce the inventory level and let the system run more effectively. In this paper, we show how kanbans can be implemented in a disassembly line setting. A disassembly line is quite different from an assembly line in terms of material movement, demand arrival and inventory level fluctuation. For example, not only can the demand arrive at the last station, it can also arrive at any of the other stations in the system. In fact, there are many other complicating matters that need to be considered to implement the concept of kanbans in such an environment. We discuss these complications and introduce the concept of a multi-kanban mechanism to cope with this environment. An example is presented to illustrate the concept.
We present a generalized stochastic Petri net (GSPN) model of a single-process, single-product, manufacturing facility with a supply management system for multiple vendors controlled by supplier kanbans. We consider a scenario with two vendors and obtain numerical results using the GSPN model. Its properties are investigated and the effect of the vendors’ activities on the system are discussed.
In this paper we demonstrate how kanbans can be used in a disassembly line setting. Disassembly line is quite different from assembly line in terms of material movement, demand arrival and inventory level fluctuation. We discuss these differences and introduce the concept of a multi-kanban mechanism to cope with them. With the help of a simulation model (developed using the ARENA® software), we show that the modified kanban mechanism is in fact an effective tool for a disassembly line setting. An example is presented to illustrate the concept.
This paper presents the effect of parameters such as demand, acquisition and sale prices of second-hand market options in reverse logistics. We analyze the secondhand market by building an integer-programming model and applying multiple degree parametric analysis to find out the critical parameters for such a reverse logistics alternative.
Disassembly activities take place in various recovery operations including remanufacturing, recycling, and disposal. The disassembly line is the best choice for automated disassembly of returned products, a feature that will be become crucial in the future. It is, therefore, important that the disassembly line be designed and balanced so that it works as efficiently as possible. However, finding the optimal balance is computationally intensive with exhaustive search quickly becoming prohibitively large, even for relatively small products. In this paper, we solve the disassembly line balancing problem using combinatorial optimization techniques, which are instrumental in obtaining near-optimal solutions to problems with intractably large solution spaces. Specifically, a genetic algorithm is presented for obtaining optimal or near-optimal solutions for disassembly line balancing problems. An example is presented to illustrate the implementation of the methodology.
Strategic planning of a distribution network is one of the most challenging aspects of reverse logistics. To effectively satisfy drivers such as profitability, environmental regulations and asset recovery, only the recovery facilities that have the potential to efficiently reprocess used-products must be considered in the reverse distribution network design. Due to uncertainties in supply, quality and reprocessing times of used-products, traditional forward logistics approaches to identify potential manufacturing facilities are not appropriate for direct adoption in reverse logistics. This paper proposes a mathematical programming approach, taking the above uncertainties into account, to effectively select potential facilities from a set of candidate recovery facilities. Application of the approach is detailed through an illustrative example.
In this paper, we investigate the effect of different dispatching rules and the vehicle assignment schemes applied to a returnable packaging system in order to provide superior customer satisfaction. The investigation is conducted by a simulation approach, using a commercial software ARENA®. The simulation results show that different combinations of dispatching rules and the vehicle assignment schemes affect the customer satisfaction levels differently.
We present a Markov model to evaluate a product recovery system with stochastic variability stemming from customer demand, recovery rate and disposal rate. The model is composed of the states that denote the number of products in inventory, the transition probabilities between states and the costs associated with the transitions. Using this model, we can calculate the total expected cost per period. An example is considered to illustrate the implementation of the methodology.
Rising fuel prices and concern over emissions are prompting automakers and legislators to introduce and evaluate "clean vehicles" throughout the United States. Hybrid electric vehicles (HEVs) are now on the roads, electric vehicles (EVs) have been test marketed, and niche vehicles such as high-fuel-economy microcars are being considered for introduction. As these vehicles proliferate and mature, they will eventually reach their end of life (EOL). In the United States, an extensive recycling infrastructure exists for conventional, internal combustion engine (ICE) vehicles. Its primary constituents are the disassembler and the shredder. These industries, as well as battery recyclers, are expected to play integral roles in the EOL processing of clean vehicles.
A model of the automobile-recycling infrastructure and goal programming techniques are used to assess the materials streams and process profitabilities for several different clean vehicles. Two-seat EVs with lead-acid or NiMH batteries are compared with two- and four-seat HEVs and microcars. Changes to the nonferrous content in the vehicle bodies are explored and compared for the effect on processing profitability. Despite limitations associated with the linearity of goal programming techniques, application of this tool can still provide informative first-order results. Results indicate that although these clean vehicles may not garner the same profit levels as conventional ICE vehicles, they are profitable to process if there are markets for parts and if there are sufficient quantities of nonferrous materials.
Strategic planning of a supply chain network is one of the most challenging aspects of reverse logistics. To effectively satisfy drivers such as profitability, environmental regula-tions and asset recovery, only the most economical used-products must be reprocessed in only the recovery facilities that have the potential to efficiently reprocess them. Due to uncertainties in supply, quality and reprocessing times of used-products, the cost-benefit function in the literature that selects the most economical product to reprocess from a set of used-products is not appropriate for direct adop-tion. Moreover, due to the same uncertainties, any tradi-tional forward supply chain approach to identify potential manufacturing facilities cannot be employed to identify potential recovery facilities. This paper proposes a three-phase mathematical programming approach, taking the above uncertainties into account, to completely design a reverse supply chain network. Application of the approach is detailed through an illustrative example in each phase.
This paper presents a novel algorithm for mesh alignment - a key step in the process of converting a physical object into its digital model. The proposed algorithm starts with meshes created by triangulation of the point cloud that results from scanning an object. The algorithm relies on the Gaussian curvature of the points on the object’s surface. The Gaussian curvatures of points on a pair of meshes (to be aligned together) are compared to find corresponding points on the pair. If the difference between the Gaussian curvatures of any two points on the mesh-pair is less than a predetermined threshold, then those two points are accepted as the corresponding points on the mesh-pair. The rotation and translation values between the mesh-pair are found by using a set of corresponding points located on them. The registration (or alignment) error is computed to check the accuracy of the mesh alignment. When the registration error becomes acceptably small, the registration process is said to reach convergence. The performance of the algorithm is studied on a test case related to a reverse engineering application. Both graphical display and extremely small registration errors indicate that the algorithm provides very good alignment results on the test model.
Often in practice, manufacturing companies involved in production of new products, also carry out collection and re-processing of used products. While environmental-consciousness is a social factor that has become an obligation to the companies in the production of new products due to governmental regulations and public perspective on environmental issues, potentiality of the companies to re-process used products is an economic factor that directly affects the profitability of the companies. Although many papers in the literature deal with performance evaluation of manufacturing companies, none of them address these two factors. To this end, a TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) approach, which evaluates manufacturing companies in terms of both environmental-consciousness and potentiality, is proposed. Furthermore, since most of the criteria that fall under these two factors are intangible, triangular fuzzy numbers (TFNs) are employed to rate them in the evaluation process. A numerical example demonstrates the feasibility of the proposed method.
Remanufacturing, recycling, and disposal recovery operations require the performance of disassembly activities. The disassembly line is the best choice for automated disassembly of returned products, however, finding the optimal balance is computationally intensive with exhaustive search quickly becoming prohibitively large. In this paper, a greedy algorithm is presented for obtaining optimal or near-optimal solutions to the disassembly line balancing problem. The greedy algorithm is a first-fit decreasing algorithm further enhanced to preserve precedence relationships. The algorithm seeks to minimize the number of workstations while accounting for hazardous and high demand components. A hill-climbing heuristic is then developed to balance the part removal sequence. Examples are considered to illustrate the methodology. The conclusions drawn from the study include the consistent generation of optimal or near-optimal solutions, the ability to preserve precedence, the speed of the algorithm and its practicality due to the ease of implementation.
The latest enhancements in industrial technologies, especially the ones in electronics industry, have provided organizations with the ability to manufacture faster and more economical products. This fact, coupled with the growing interest and demand for the latest technology, have led electronic equipment manufacturers to start producing "high-tech" and "personalized" products at an increasing rate. This has led to a high rate of obsolescence for electronic products worldwide, even though the majority of these "obsolete" products still function. In this paper, we investigate a product recovery facility where the end-of-life (EOL) products are taken back from the last users and are brought into the facility for processing. We assume that there are multiple types of EOL products and that a combination of these can be disassembled to provide for a sufficient number of demanded components and materials. We then present a data envelopment analysis (DEA) algorithm to determine the number and types of the EOL products that will be required to fulfill the demand. A numerical example is presented to demonstrate the functionality of the methodology.
In recent years, the continuous growth in consumer waste and dwindling natural resources has seriously threatened the environment. Realizing this, several countries have passed regulations that force manufacturers not only to manufacture environmentally conscious products, but also to take back their used products from consumers so that the components and materials recovered from the products may be reused and/or recycled. Disassembly plays an important role in product recovery. A disassembly line is perhaps the most suitable setting for disassembly of products in large quantities. Because a disassembly line has a tendency to generate excessive inventory, employing a kanban system can reduce the inventory level and let the system run more efficiently. A disassembly line is quite different from an assembly line. For example, not only can the demand arrive at the last station, it can also arrive at any of the other stations in the system. The demand for a component on the disassembly line could fluctuate widely. In fact, there are many other complicating matters that need to be considered to implement the concept of kanbans in such an environment. In this paper, we discuss the complications that are unique to a disassembly line. We discuss the complications in utilizing the conventional production control mechanisms in a disassembly line setting. We then show how to overcome them by implementing kanbans in a disassembly line setting with demand fluctuation and introduce the concept of multi-kanban mechanism. We demonstrate its effectiveness using a simulation model. An example is presented to illustrate the concept.
This paper deals with the cost management problem of a remanufacturing system with stochastic variability in the demand rate, the remanufacturing rate and the discard rate. We consider two types of inventories. One is the actual product inventory in the factory while the other is the virtual inventory that is still in use by the consumers. The state of the remanufacturing system is defined by considering the levels of both inventories. The cost function is composed of various costs such as the holding cost, backlogging cost and other manufacturing costs. We obtain the optimal production policy that minimizes the expected average cost per period. Numerical results provide insights on the effects of the various costs on the optimal policy.
It has become common for manufacturing facilities involved in production of new products to also carry out collection and re-processing of used products. While environmental consciousness has become an obligation to the facilities in the production of new products due to governmental regulations and public perspective on environmental issues, potentiality of the facilities to re-process used products directly affects the profitability of the facilities. Although many papers in the literature deal with performance evaluation of facilities, none of them address these two factors. To this end, a TOPSIS (Technique for Order Preference by Similarity to an Ideal Solution) approach, which evaluates production facilities in terms of both environmental-consciousness and potentiality, is proposed. Furthermore, since most of the criteria that fall under these two factors are intangible, triangular fuzzy numbers (TFNs) are employed to rate them in the evaluation process. A numerical example demonstrates the feasibility of the proposed method.
The cost-benefit analysis of data associated with re-processing of used products often involves the uncertainty feature of cash-flow modeling. The data is not objective because of uncertainties in supply, quality and disassembly times of used products. Hence, decision-makers must rely on "fuzzy" data for analysis. The same parties that are involved in the forward supply chain often carry out the collection and re-processing of used products. It is therefore important that the cost-benefit analysis takes the data of both new products and used products into account. In this paper, a fuzzy cost-benefit function is proposed that is used to perform a multi-criteria economic analysis to select the most economical products to process in a closed-loop supply chain. Application of the function is detailed through an illustrative example.
Although there are many quantitative models in the literature to design a reverse supply chain, every model assumes that all the recovery facilities that are engaged in the supply chain have enough potential to efficiently re-process the incoming used products. Motivated by the risk of re-processing used products in facilities of insufficient potentiality, this paper proposes a method to identify potential facilities in a set of candidate recovery facilities operating in a region where a reverse supply chain is to be established. In this paper, the problem is solved using a newly developed method called physical programming. The most significant advantage of using physical programming is that it allows a decision maker to express his preferences for values of criteria (for comparing the alternatives), not in the traditional form of weights but in terms of ranges of different degrees of desirability, such as ideal range, desirable range, highly desirable range, undesirable range, and unacceptable range. A numerical example is considered to illustrate the proposed method.
Collectors of discarded products seldom know when those products were bought and why they are discarded. Also, the products do not indicate their remaining life periods. So, it is difficult to decide if it is "sensible" to repair (if necessary) a particular product for subsequent sale on the second-hand market or to disassemble it partially or completely for subsequent remanufacture and/or recycle. To this end, we build an expert system using Bayesian updating process and fuzzy set theory, to aid such decision-making. A numerical example demonstrates the building approach.
Returnable transport packaging plays an important role in facilitating the transfer of a large volume of products in a close-loop distribution network. To make effective use of returnable transport packaging, vehicle dispatching strategies are crucial. With an appropriate vehicle dispatching strategy, for example, a fast turnover time and a short waiting time for packaging dispatch can be achieved. However, there are some factors that directly influence vehicle dispatching strategies. These factors include the arrival demand fluctuations, the availability of serving vehicles, and the geographic proximity of the facility to the customer’s locations. In this study, authors investigate the effect of these factors on vehicle dispatching strategies for transport packaging by using a simulation modeling approach. This paper reports different performance outcomes obtained through various test cases.
The main thrust of returnable packaging these days is to provide logistical services through transportation and distribution of products and be environmentally friendly. Returnable packaging and reverse logistics concepts have converged to mitigate the adverse effect of packaging materials entering the solid waste stream. Returnable packaging must be designed by considering the trade-offs between costs and environmental impact to satisfy manufacturers and environmentalists alike. The cost of returnable packaging entails such items as materials, manufacturing, collection, storage and disposal. Environmental impacts are explicitly linked with solid waste, air pollution, and water pollution. This paper presents a multi-criteria evaluation technique to assist decision-makers for evaluating the trade-offs in costs and environmental impact during the returnable packaging design process. The proposed evaluation technique involves a combination of multiple objective integer linear programming and analytic hierarchy process. A numerical example is used to illustrate the methodology.
Disassembly activities are an important part of product recovery operations. The disassembly line is the best choice for automated disassembly of returned products. However, finding the optimal balance for a disassembly line is computationally intensive with exhaustive search quickly becoming prohibitively large. In this paper, a greedy algorithm is presented for obtaining optimal or near-optimal solutions to the disassembly line-balancing problem. The greedy algorithm is a first-fit decreasing algorithm further enhanced to preserve precedence relationships. The algorithm seeks to minimize the number of workstations while addressing hazardous and high demand components. A two optimal algorithm is then developed to balance the part removal sequence and attempt to further reduce the total number of workstations. Examples are considered to illustrate the methodology. The conclusions drawn from the study include the consistent generation of optimal or near-optimal solutions, the ability to preserve precedence, the speed of the algorithms and their practicality due to the ease of implementation.
Taking a multi-resolution approach, this research work proposes an effective algorithm for aligning a pair of scans obtained by scanning an object's surface from two adjacent views. This algorithm first encases each scan in the pair with an array of cubes of equal and fixed size. For each scan in the pair a surrogate scan is created by the centroids of the cubes that encase the scan. The Gaussian curvatures of points across the surrogate scan pair are compared to find the surrogate corresponding points. If the difference between the Gaussian curvatures of any two points on the surrogate scan pair is less than a predetermined threshold, then those two points are accepted as a pair of surrogate corresponding points. The rotation and translation values between the surrogate scan pair are determined by using a set of surrogate corresponding points. Using the same rotation and translation values the original scan pairs are aligned. The resulting registration (or alignment) error is computed to check the accuracy of the scan alignment. When the registration error becomes acceptably small, the algorithm is terminated. Otherwise the above process is continued with cubes of smaller and smaller sizes until the algorithm is terminated. However at each finer resolution the search space for finding the surrogate corresponding points is restricted to the regions in the neighborhood of the surrogate points that were at found at the preceding coarser level. The surrogate corresponding points, as the resolution becomes finer and finer, converge to the true corresponding points on the original scans. This approach offers three main benefits: it improves the chances of finding the true corresponding points on the scans, minimize the adverse effects of noise in the scans, and reduce the computational load for finding the corresponding points.
Smart sensors and their networking technology when applied in manufacturing environment for monitoring, diagnostics, and control and for data/information collection could dwarf all the advances made so far by the manufacturing community through traditional sensors. Smart sensors can significantly contribute to improving automation and reliability through high sensitivity, self-calibration and compensation of non-linearity, low-power operation, digital pre-processed output, self-checking and diagnostic modes, and compatibility with computers and other subsystem blocks. There is a huge gulf between the existing models of manufacturing systems and the computational models that are required to correctly characterize manufacturing systems integrated with smart sensor networks. This paper proposes a multi-agent model for S 2 IM system. The agent characteristics and the expected model behavior are presented.
The disassembly line is the best choice for automated disassembly of returned products. However, finding the optimal balance for a disassembly line is computationally intensive with exhaustive search quickly becoming prohibitively large. Metaheuristic techniques provide a general algorithmic framework that can be applied to this optimization problem. Although metaheuristics show promise in solving this complex problem, challenges exist in the variety of evaluation criteria available, a lack of disassembly-specific data sets for metaheuristic testing and a lack of performance analysis tools. In this paper, a balance performance measure is reviewed along with a size-independent a priori data set and graphical analysis tools.
Evaluating the performance of a recycling program is equivalent to evaluating how well the program is driving the public to participate in the program. Studies are conducted in numerous cities around the world, in order to assess the level of participation of the public in their respective recycling programs. The officials of each program painstakingly approach many homes in the city, with questions regarding how convenient the program is to them and how the program can be improved. This paper lists some important drivers of public participation, and proposes a fuzzy TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) approach to evaluate recycling programs with respect to those drivers. Application of the fuzzy TOPSIS approach is detailed through a numerical example.
We propose a neural network approach to evaluate the success potential of a collection center of interest that is being considered for inclusion in a reverse supply chain, using the available linguistic data of collection centers that already exist in the reverse supply chain. The approach is carried out in four phases, as follows. In phase I, we identify the criteria for evaluating the collection center of interest, by each group participating in the reverse supply chain, viz. consumers, governments and executives. Then, in phase II, we use fuzzy ratings of already existing collection centers to construct a neural network that gives impacts of criteria identified for each group in phase I. In phase III, using the impacts obtained in phase II, we employ a fuzzy TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) approach to obtain the overall rating of the collection center of interest, as evaluated by each group. Finally, in phase IV, we employ Borda’s choice rule to calculate the maximized consensus rating, i.e., success potential of the collection center of interest. An example is considered to illustrate the approach.
In this paper, we propose an application of a pull type production control mechanism for a disassembly line setting. We discuss complications and justify the use of pull type mechanism in a disassembly line. We introduce a multi-kanban mechanism for a multi-product disassembly line where demand for components can arrive at any level. We define the kanban routing rules to minimize the system’s inventory while maintaining a customer service level comparable to a push system. We suggest a method of determining the proper number of base kanbans and demonstrate its effectiveness by using a simulation model and implementing it in a case example.
Product disassembly takes place in remanufacturing, recycling, and disposal. The disassembly line is the best choice for automated disassembly, so it is essential that it be designed and balanced to work efficiently. The multi-objective disassembly line balancing problem (DLBP) seeks to find a disassembly sequence which: provides a feasible disassembly sequence, minimizes the number of workstations, minimizes idle time, and balances the line (ensure similar idle times at each workstation) as well as addressing other, disassembly-specific concerns. However, finding the optimal balance is computationally intensive due to factorial growth, with exhaustive search quickly becoming prohibitively large. In this paper, an ant colony optimization metaheuristic is presented for obtaining optimal or near-optimal solutions to the DLBP. Examples are considered to illustrate implementation of the methodology. Conclusions drawn include the consistent generation of near-optimal solutions, the ability to preserve precedence, the superior speed of the metaheuristic, and its practicality due to its ease of implementation.
Although many quantitative models have been reported in the literature for designing a supply chain, none of them address issues that are crucial for the supply chain to successfully operate in a closed- loop. To that end, we examine the following: (i) selection of economical products to process, (ii) identification of efficient production facilities in the region where the supply chain is to be designed, (iii) optimal transportation of products across the supply chain, (iv) prediction of success potentials of collection centers that are being considered for inclusion in the supply chain, (v) sale of used products on potential second-hand markets, (vi) evaluation of marketing method - especially in the reverse supply chain, and (vii) involvement of consumers and local government, along with company executives, in decision- making. Possible strategies to approach these issues are also suggested.
In this paper, a multi-Kanban model for disassembly line with multi-product and multiple demands is presented. We define the Kanban routing rules to minimize the system’s inventory while maintaining a customer service level comparable to a push system. We determine the number of base Kanban level by taking product arrival rate, demand arrival rate, and disassembly times into consideration. We provide a numerical example to illustrate the methodology and obtain results using simulation. The results demonstrate the superiority of the proposed method in controlling the system’s inventory over the push system.
For a prospective reverse supply chain to operate efficiently, the designing of that chain must involve selection of collection centers and recovery facilities that have sufficient success potentials. These success potentials depend heavily on the participation (in the reverse supply chain) of three important groups that have multiple, conflicting, and incommensurate goals. Therefore, the potentials must be evaluated based on the maximized consensus among the three groups, viz., (i) Consumers whose primary concern is convenience, (ii) Local government officials whose primary concern is environmental consciousness, and (iii) Supply chain company executives whose primary concern is profit. In this paper, we propose a three-phase multi-criteria group approach to select collection centers as well as recovery facilities, of sufficient success potentials. In the first phase of the approach, we identify important criteria for evaluation of the alternatives (collection centers as well as recovery facilities) for each of the above three groups. In the second phase, we give weights to the criteria of each group using the Eigen vector method, and then, employ the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) to find the success potential of each alternative, as evaluated by that group. Then, in the third and final phase, we use Borda’s choice rule that, for each alternative, combines individual success potentials into a group success potential or maximized consensus ranking. We also employ a GIS-based application, MapLand, to "map" the results obtained in the second and the third phases of our approach.
A two-phase mathematical programming approach to efficiently design a reverse supply chain is proposed, as follows: in phase I, a fuzzy cost-benefit function is formulated to perform a multi-criteria economic analysis for selecting the most economical product to re-process, from a set of candidate used products; in phase II, an integer goal programming model that not only identifies potential recovery facilities but also leads to transportation of the right mix and quantities of products (used as well as re-processed) across the supply chain, is formulated. Since the success of a reverse supply chain depends on its marketing strategy as well as on its design, it is important that the planned marketing strategy be evaluated with respect to drivers of public participation in the supply chain (more participation of the public implies more effectiveness of the marketing strategy), before actually implementing the strategy. To this end, we identify all the important drivers of public participation, and propose a fuzzy TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) approach to evaluate the marketing strategy of a reverse supply chain with respect to those drivers.
The disassembly line is the best choice for automated disassembly of returned products. However, finding the optimal balance is computationally intensive with exhaustive search quickly becoming prohibitively large. In this paper, a new general-purpose heuristic algorithm is proposed for obtaining near-optimal solutions to the disassembly line-balancing problem. An example is considered to illustrate the methodology and demonstrate the metaheuristics performance. The conclusions drawn from the study include the consistent generation of near-optimal solutions, the ability to preserve precedence, the superior speed of the metaheuristic and its practicality due to the ease of implementation.
Although many quantitative models have been reported in the literature for designing (also called strategic planning) a reverse supply chain, every model assumes that all the facilities that are engaged in the reverse supply chain are profited by re-processing economical “used products” or cores. Further, every model also assumes that the facilities have sufficient potential to efficiently re-process the incoming cores. To these ends, this paper proposes a three-phase linear physical programming approach for designing a reverse supply chain. In phase I, an approach to select economical products to re-process from a set of candidate cores is proposed. In phase II, an approach to identify potential recovery facilities in a region where the reverse supply chain is to be designed, is proposed. Finally, in phase III, a single time-period transshipment model is solved to achieve transfer of the right mix and quantities of products (used as well as re-processed) across the supply chain. Linear physical programming is a newly developed method whose most significant advantage is that it allows a decision-maker to express his preferences for values of criteria for decision making, not in the traditional form of weights but in terms of ranges of different degrees of desirability.
The disassembly line is the best choice for automated disassembly of returned products. However, finding the optimal balance of the line is computationally intensive with exhaustive search quickly becoming prohibitively large. In this paper, a greedy algorithm is initially presented for obtaining solutions to the disassembly line balancing problem. The First-Fit- Decreasing based algorithm seeks to minimize the number of workstations while accounting for hazardous and high-demand components. A hill-climbing heuristic, AEHC, is then used to improve the balance of the part removal sequence. These are compared with an implementation of the H-K metaheuristic via a case study from the literature.
Selection of an optimal disassembly sequence is essential for the efficient processing of a product at the end of its life. Disassembly sequences are listings of disassembly actions (such as the separation of an assembly into two or more subassemblies, or removing one or more connections between components). Disassembly takes place in remanufacturing, recycling, and disposal with a disassembly line being the best choice for automation. In this paper, the disassembly sequencing problem is solved for a cell phone case on a disassembly line, seeking a sequence which is feasible, minimizes the number of workstations (and hence idle times), provides for early removal of high demand/value parts, provides the removal of parts that lead to the access of greatest number of still-installed parts, and early removal of hazardous parts as well as for the grouping of parts for removal having identical part removal directions. Since finding the optimal sequence is computationally intensive due to factorial growth, a heuristic method is used taking into account various disassembly-specific matters. Using the experimentally determined precedence relationships and task times of a real-world cell phone, a MATLAB program is designed and a sequencing solution is generated. Finally, Design for Disassembly (DFD) improvements are recommended with respect to environmentally conscious manufacturing.
Disassembly takes place in remanufacturing, recycling, and disposal with a line being the best choice for automation. The disassembly line balancing problem seeks a sequence which: minimizes workstations, ensures similar idle times, and is feasible. Finding the optimal balance is computationally intensive due to factorial growth. Combinatorial optimization methods hold promise for providing solutions to the disassembly line balancing problem, which is proven to belong to the class of NP-complete problems. Ant colony optimization, genetic algorithm, and H-K metaheuristics are presented and compared along with a greedy/hill-climbing heuristic hybrid. A numerical study is performed to illustrate the implementation and compare performance. Conclusions drawn include the consistent generation of optimal or nearoptimal solutions, the ability to preserve precedence, the speed of the techniques, and their practicality due to ease of implementation.
The use of personal computers (PCs) continues to increase every year. According to a 1999 figure, 50 percent of all US households owned PCs, a figure that continues to rise every year. With continuous development of sophisticated software, PCs are becoming increasingly powerful. In addition, the price of a PC continues to steadily decline. Furthermore, the typical life of a PC in the workplace is approximately two to three years while in the home it is three to five years. As these PCs become obsolete, they are replaced and the old PCs are disposed of. It is estimated that between 14 and 20 million PCs are retired annually in the US. While 20 to 30% of the units may be resold, the others are discarded. These discards represent a significant potential source of lead for the waste stream. In some communities, waste cathode ray tubes (CRTs) represent the second largest source of lead in the waste stream after vehicular lead acid batteries. PCs are, therefore, not suitable for dumping in landfills. Besides, several components of a PC can be reused and then there are other valuable materials that can also be harvested. And with the advent of product stewardship, product recovery is the best solution for manufacturers. Disassembly line is perhaps the most suitable set up for disassembling PCs. However, planning and scheduling of disassembly on a disassembly line is complicated. In this paper, we discuss some of the complications including product arrival, demand arrival, inventory fluctuation and production control mechanisms. We then show how to overcome them by implementing a multi-kanban mechanism in the PC disassembly line setting. The multi-kanban mechanism relies on dynamic routing of kanbans according to the state of the system. We investigate the multi-kanban mechanism using simulation and demonstrate that this mechanism is superior to the traditional push system in terms of controlling the system’s inventory while maintaining a decent customer service level.
In this paper we present a dynamic kanban (pull) system specifically developed for disassembly lines. This type of kanban system is much more complex than the traditional kanban system used in assembly lines. For instance, unlike the assembly line where the external demand occurs only at the last station, the demands in the disassembly case also occur at any of the intermittent stations. The reason is that as a product moves on the disassembly line, various parts are disassembled at every station and accumulated at that station. Therefore, there are as many demand sources as there are number of parts. We consider a case example involving the end-of-life products. Based on the precedence relationships and other criteria such as hazardous properties of the parts, we balance the disassembly line. The results of the disassembly line-balancing problem (DLBP) are used as input to the proposed dynamic kanban system for disassembly line (DKSDL). We compare the performance of the DKSDL to the modified kanban system for disassembly line (MKSDL), which was previously introduced by the authors. We show, via simulation, that the DKSDL is far superior to MKSDL considered.
The designing of a reverse supply chain must involve selection of collection centers and recovery facilities that have sufficient success potentials. These success potentials depend heavily on the participation of the following three important groups who have multiple, conflicting, and incommensurate criteria for evaluation, and so, the potentials must be evaluated based on the maximized consensus among those groups: (i) Consumers (whose primary concern is convenience), (ii) Local government officials (whose primary concern is environmental consciousness), and (iii) Supply chain company executives (whose primary concern is profit). In this paper, we propose a three-phase multi-criteria group approach to select collection centers as well as recovery facilities, of sufficient success potentials. In the first phase of the approach, we identify important criteria for evaluation of the alternatives (collection centers as well as recovery facilities) by each of the above three groups. In the second phase, we give weights to the criteria of each group using the eigen vector method, and then, employ the TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) to find the success potential of each alternative, as evaluated by that group. Then, in the third and final phase, we use Borda’s choice rule that, for each alternative, combines individual success potentials into a group success potential.
It is difficult to obtain information regarding compositions and remaining life periods of used products. Hence, they often undergo partial or complete disassembly for subsequent re-processing (remanufacturing and/or recycling). However, researchers are now studying sensor embedded products (SEPs), the composition and remaining life of which can be obtained at the end of their use from sensors. This paper addresses decision-making regarding the futurity of an SEP at its end of use: whether to disassemble it for subsequent recycling/remanufacturing or to repair it for subsequent sale on a second-hand market. We identify some important factors that must be considered before making a decision. Using a numerical example, we propose a simple approach that employs Bayesian updating and fuzzy set theory to aid the decision-making process.
In this paper, we consider the problem of determining the optimal number of returned products to disassemble to fulfill the demand for a specified number of parts. This is known as the disassembly-to-order (DTO) problem. The deterministic yield version of this problem has been addressed in the literature. Recently, the stochastic yield version of this problem with a single objective has also been reported in the literature. In this paper, we extend the methodology to include multiple objectives. To this end, we model the DTO problem using integer goal programming. The stochastic problem is solved by transforming it into its deterministic equivalent problem. This conversion is accomplished by considering the specific structures of the products with one core and one part (“one-to-one structure”) and apply it to handle the products with one core and multiple parts (“one-to-many structure”). For these special cases it is possible to solve the stochastic problem analytically so that valuable insights can be gained by comparing the stochastic and deterministic solutions. This will help us to determine effective deterministic yield equivalents. We present a case example to illustrate the methodology.
The driving forces behind product take-back and green manufacturing are well established. The two main product end-of-life options are reuse/remanufacturing and recycling. For either option, all take-back units are treated equally because no information that tracks the conditions of a product during its useful life is available. For example, all expired PCs are treated equally; no distinction can be made about which units still have healthy hard disks. This paper discusses sensor-based monitoring and prognostic methodologies for tracking the condition of products while being used by customers and timely and targeted servicing, smart and selective disassembling and refurbishing of products with known (long) remaining lives. The paper also discusses the added benefits to product manufacturers when the time comes to redesign their products. The real-time field data on service and utilization of products are communicated to manufacturers’ headquarters for further analysis.
Packaging material selection (PMS) problems have always been important to packaging designers and engineers. Not only does the selection of packaging material determine the costs and the environmental impacts of packaging, but also influences packaging physical characteristics and associated manufacturing methods. In order to reduce economic and environmental impacts, one has to take a holistic approach to packaging material selection by considering material effects throughout the packaging life cycle. To evaluate economic costs and environmental impacts both quantitative factors and subjective criteria play an important role in the packaging design. In the present work, fuzzy set theory is used for representing and manipulating the vague and subjective descriptions of packaging performance and design attributes. Further a genetic algorithm based approach is used for addressing the packaging material selection problem through multiple criteria decision-making. The overall approach comprises of two phases. In the first phase, fuzzy set theory is used for the linguistic transformation of performance attributes into numerical values. It results in a decision matrix that contains crisp scores. Also in this phase, a weight is assigned to each sub-criterion to show its importance compared to others. In the second phase, a GA is used to globally search for near-optimal or optimal design solutions. The implementation of the proposed methodology is illustrated through a numerical example.
Recently work has been done about simulation and analysis of manufacturing operations combined with the communications network it operates with. However no such methodology exists in disassembly line context. In this paper, we attempt to enhance their approach and apply it for a disassembly line. To this end, we look at a disassembly line system as a combination of physical processing (performed by machines) and information processing (performed by computer systems), and formulate a model to analyze the system behavior and to obtain an optimal or near optimal solution that would maximize the system performance by minimizing the risk of down time due to network capacity related problems. A case example is presented to demonstrate the feasibility of the model’s implementation.
This research investigates the advantages offered by embedded sensors for cost-effective and environmentally benign product life cycle management for desktop computers. During their use by customers as well as at the end of their lives, Sensor Embedded Computers (SECs) by virtue of sensors embedded in them generate data and information pertaining to the conditions and remaining lives of important components such as hard-drive, motherboard, and power supply unit. A computer monitoring framework is proposed to provide more customer comfort, reduce repair costs and increase the effectiveness of current disassembly practices. The framework consists of SECs, remote monitoring center (RMC), repair/service, disassembly, and disposal centers. The RMC collects dynamic data/information generated by sensors during computer usage as well as static data/information from the original equipment manufacturers (OEMs). The RMC forwards this data/information to the repair/service, disassembly, and disposal centers on need-basis. The knowledge about the condition and remaining life of computer components can be advantageously used for planning repair/service and disassembly operations as well as for building refurbished computers with known expected lives. Simulation model of the framework is built and is evaluated in terms of the following performance measures: average downtime of a computer, average repair/service cost of a computer, average disassembly cost of a computer, and average life cycle cost of a computer. Test results show that embedding sensors in computers provides a definite advantage over conventional computers in terms of the performance measures.
Remanufacturing operations involved with highly uncertain recovery rate of used products, subassemblies and parts that complicate the planning and control of the process. In this paper, we develop a comprehensive procedure that determines the optimal input quantities at each stage of the remanufacturing operations in which recovery rates at each stage of the process are stochastic. We model the remanufacturing system as an open queueing network and use the decomposition principle and expansion methodology to analyze it. Each station in the system is subject to breakdown and has a finite buffer capacity. Repair times, breakdown times and service times follow exponential distributions. Optimization is done on the system’s expected total cost using a dynamic programming (DP) algorithm. A numerical example is presented to show the applicability of the model.
Disassembly takes place in remanufacturing, recycling, and disposal with a line being the best choice for automation. The disassembly line balancing problem seeks a sequence which: is feasible, minimizes workstations, and ensures similar idle times, as well as other, disassembly-specific concerns. Finding the optimal balance is computationally intensive due to factorial growth. Ant colony optimization and genetic algorithm metaheuristics are presented and compared along with an exa mple to illustrate the implementation. Conclusions drawn include the consistent generation of optimal or near-optimal solutions, the ability to preserve precedence, the speed of the metaheuristics, and their practicality due to ease of implementation.
We believe that for some end-of-use products, it might make more “sense” to make necessary repairs to the products and sell them on second-hand markets than to disassemble them for subsequent re-processing (remanufacturing and/or recycling). To this end, we propose an approach that uses Bayesian updating process and fuzzy set theory, to decide if it is “sensible” to repair an improperly functioning end-of-use product of interest for subsequent sale on a second-hand market. Furthermore, we employ the fuzzy Quality Function Deployment (QFD) process as well as the method of total preferences, to select the most potential market to sell the repaired end-of-use product on, from a set of candidate second-hand markets.
Increasing environmental concerns during the last decade have caused many governments to persuade manufacturers to take back used products so that the components and materials recovered from the products could be reused and/or recycled. The disassembly process is the first step for recovering components and materials. We discuss some of the complications in planning and scheduling of disassembly on a disassembly line. We show how to overcome them by implementing a multi-kanban mechanism in the disassembly line setting. We then investigate the multi-kanban mechanism using simulation and demonstrate that in the multi-product, multi-demand environment, such mechanism performs superior to the traditional push system.
Strategic planning of a supply chain network is one of the most challenging elements of managing reverse logistics. To effectively satisfy drivers such as profitability, environmental regulations and asset recovery, only the most economical used-products must be re-processed in only the recovery facilities that have the potential to efficiently re-process them. Due to uncertainties in supply, quality and re-processing times of used-products, the cost–benefit function used in the literature to select the most economical product to re-process from a set of used-products, is not appropriate for direct adoption. Moreover, due to the same uncertainties, any traditional forward supply chain approach to identify potential manufacturing facilities cannot be employed to identify potential recovery facilities. This paper proposes a three-phase mathematical programming approach, taking the above uncertainties into account, for strategic planning of a reverse supply chain network. Application of the approach is detailed through illustrative examples.
We present a near optimal buffer allocation plan (NOBAP) specifically developed for a remanufacturing cell with finite buffers and unreliable servers. A remanufacturing cell is a self-reliant entity within which a variety of activities such as disassembly, inspection, material processing, remanufacturing, assembly and transportation are performed. The remanufacturing cell considered in this paper consists of three modules, viz. the disassembly and testing module for returned products, the disposition module for non-reusable returns and the remanufacturing module. We propose an algorithm that uses an open queueing network, decomposition principle and expansion methodology to analyze the remanufacturing cell. The buffer allocation algorithm distributes a given number of available buffer slots among the various stations (across the various modules) to optimize the cell’s performance. The algorithm has been rigorously tested for both balanced and unbalanced cells. The results show that the performance of the algorithm is consistent, robust and produces excellent results in a variety of experimental conditions.
Industry has grown to recognize the value of disassembly processes across a wide range of products. Increasing legislation that may soon require mandatory recycling of many post-consumed goods and a desire to develop more environmentally benign end-of-life processes has fueled research into this concept. Traditionally, disassembly has been viewed as the reverse of assembly; however, a novel view considers just the opposite, leading to a more optimized disassembly process.
Disassembly Modeling for Assembly, Maintenance, Reuse, and Recycling presents this approach in the context of the entire product life cycle. The book examines disassembly on the intermediate level, incorporating design for disassembly, concurrent design, and reverse logistics. In this first text to supply a comprehensive discussion of the theories and methodologies associated with this approach, the authors incorporate real world case examples to explore the three main areas of application of the theory: assembly optimization, maintenance and repair, and end-of-life processing.
This is a timely resource for companies that wish to enact environmentally conscious systems efficiently. With an analysis of associated costs, system design requirements, advantages, and expected results, this is also an indispensable tool for researchers, mechanical and industrial engineers, and professionals involved in concurrent design.
(Table of Contents), (Author Index), (Subject Index), (Order).
Disassembly takes place in remanufacturing, recycling and disposal, with a flow shop being the best choice for automation. The disassembly line balancing problem seeks a sequence which: is feasible, minimizes workstations, and ensures similar idle times, as well as other end-of-life specific concerns. Finding the optimal balance is computationally intensive due to exponential growth. Combinatorial optimization methods hold promise for providing solutions to the problem, which is proven NP-hard. Stochastic (genetic algorithm) and deterministic (greedy/hill-climbing hybrid heuristic) methods are presented and compared. Numerical results are obtained using a recent electronic product case study.
In this paper, we describe the implementation of a modified-pull system in a disassembly line setting, particularly the one with component-discriminating demand. The many complications associated with a disassembly line make it difficult for the traditional pull system to operate efficiently in this environment. These complications include: fluctuation in demand, supply and inventory, divergent flow of materials, and difficulty in selecting end-of- life products. The multikanban system proposed in this research utilizes enhanced routing rules and other policies to find a balance between system’s efficiency and its ability to satisfy the customer’s demand. A numerical example is considered for illustrating the effectiveness of the proposed system.
Machine-tool breakdowns, both expected and unexpected, make it difficult to devise effective maintenance plans that do not adversely affect the performance of a manufacturing system. This work presents a heuristic maintenance grouping algorithm which takes the machine-tool failure information provided by smart sensors into account to come up with condition-based preventive maintenance plans that does not severely affect the manufacturing system operations. Studies on the simulation model of a flow-line manufacturing system integrated with smart sensors indicate that grouping of maintenance actions increases the uptime efficiency, boosts the production rate, reduces the maintenance cost, and decreases the production loss.
In this paper we study the effect of the sudden breakdown of material handling equipment on the performance of recently introduced MKSD (Modified Kanban System for Disassembly). We also study behavior of a newly developed JIT system for disassembly called DKSDL (Dynamic Kanban System for Disassembly Line) under the same conditions. DKSDL dynamically adjusts the number of disassembly production kanbans in or der to offset blocking and starvation caused by these factors during a disassembly cycle. We compare the overall performances of the MKSD and the DKSDL models under different scenarios. The scenario settings, approach to the solution, results and discussion of these scenarios are included.
In this paper, we consider the disassembly-to-order (DTO) problem, where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The objective is to determine the optimal numbers of returned products to disassemble so as to maximize profit and minimize costs. We model the DTO problem using a multi-criteria decision making approach. Since the condition of returned products are unknown, the yields from disassembly are considered to be stochastic. To solve the stochastic problem, we use one of two heuristic approaches (viz., one-to-one approach or one-to-many approach) that convert the problem into a deterministic equivalent. We compare the performance of the two heuristic approaches using a case example.
Communication networks play an increasingly important role in assembly and disassembly systems. Most of the works published recently to analyze such systems use simulation. However, these works have typically focused on Carrier Sense Multiple Access protocol and none of the studies address the performance of Token Ring protocol for such applications. In this paper, we attempt to fill this void. To this end, we analyze a disassembly system that is a combination of physical manifold (performed by machines) and information network (performed by computers). We develop a model to analyze the system's behavior and attempt to obtain an optimal or near optimal solution to maximize the system's performance by minimizing the risk of downtime due to network capacity related problems. A case example is considered to demonstrate the feasibility of the model’s implementation.
We develop a back-propagation neural network (BPN) to predict the life-cycle design performance for transport packaging. The BPN is constructed and trained on the packaging design attributes to detect hidden relationships among historical or pre-existing life-cycle design data to predict a new concept design through supervised learning, by minimizing the squared difference between the actual and the predicted life-cycle design performance. To this end, the designer could use the predicted life-cycle design in a trade-off analysis and concept selection for a potential packaging design. A case example is used to illustrate the methodology.
In this paper, we consider the disassembly-to-order (DTO) system, where end-of-life (EOL) products are taken back to be disassembled to satisfy demands for components and materials. The main purpose of this research is to determine the optimal number of take-back EOL products for the DTO system while trying to maximize profit and minimize costs involved in the system. In order to achieve the goals, we implement a multi-criteria decision making approach to handle such multiple objectives. Since the operating conditions of take-back EOL products are unknown and always complicate model formulation, heuristic approach which transforms stochastic version of disassembly yields into its deterministic equivalents is included in our work. In this research, we attempt to generate a DTO plan for multiple periods, where the remaining or exceeding resources in the previous period can be used in the following periods. A numerical example is cons idered to illustrate the approach.
Rapid developments in present day computer technology is rendering personal computers (PCs) life span short, thus, increasing the number of PC’s to be planned for end-of-life (EOL) processing. In this paper, we develop a Cost-Benefit model for a disassembler of EOL PC’s and a buyer of the disassembled parts of a PC in the second hand market. We carry out a sensitivity analysis of the variables that affect the profit of both the parties, such that, both parties can maximize their mutual profits. We also consider a special case where a disassembler is an original equipment manufacturer (OEM) and model a specific scenario where the OEM encourages his/her customers to bring back their existing PC’s that are relatively new, at the same time, offering brand new PC’s at a discounted price, the discount being a variable dependent upon the age of the PC the customer is bringing back. A numerical example is considered to illustrate the methodology.
In this paper, a multi-kanban mechanism is proposed for automobile disassembly. We concentrate on a service facility, which deals with a disassembly line and large volume of components and vehicles. Because customer demand often discriminates among vehicle’s makes and models, we employ multiple types of component kanbans at each disassembly workstation. We define the kanban routing rules and the component selection rules to minimize the system’s inventory while maintaining a customer service level comparable to a push system. We consider a numerical example to illustrate the methodology and obtain results using simulation. The results demonstrate the effectiveness of utilizing the multi-kanban mechanism in the automobile disassembly line.
Automation is a very important part of dis/assembly lines. Communication network is the medium and set of rules, which transfers and regulates all the necessary information flow to synchronize the automation on the line and to control each product on the line. However, very little work has been done to analyze the effects of different kinds of communication networks on the performance of the line. In this paper, we look at a disassembly line system as a combination of physical processing (performed by machines) and information processing (performed by computer systems), and formulate a model to analyze the system behavior and to obtain an optimal or near optimal solution that would maximize the system performance by minimizing the risk of down time due to network capacity related problems. We simulate this model for two major types of network protocols (CSMA/CD and Token Ring) used in the manufacturing and de-manufacturing industries. A comparison of their performances on the throughput of a TV disassembly line is included as a case study.
A few location models have been reported in the literature for strategic planning (also called designing) of reverse supply chains. In the case of discrete location models, all recovery facilities are assumed to be efficient and in the case of continuous location models, it is assumed that efficient recovery facilities were already established or can be established at the locations solved for. Also, each of these location models deals with a used product that is assumed to be economical. Evidently, though every location model realizes the importance of re-processing (recycling /remanufacturing) only economical used products in efficient recovery facilities, it does not show how to either select those used products from a set of different used products or identify those recovery facilities in a region where the reverse supply chain is to be designed. Addressing these limitations, this paper proposes a multi-phase approach for strategic planning of a reverse supply chain.
End-of-life products usually contain reusable materials and components. Disassembly can help increase material and component recovery rate. In selective disassembly of EOL products, a disassembly line is more commonly employed than other disassembly setups. Due to the uncertainty regarding the condition of EOL products, the decision whether to disassemble them has to be made impromptu which can cause delays in satisfying the demand for components. Sensors, if embedded in products, can provide information pertaining to product condition which can assist in decision making during product disassembly, thereby boosting product recovery rate, reducing labor cost, increasing line efficiency, and minimizing disassembly time. This work proposes a model of a disassembly line for sensor-embedded products (SEPs) with pull mechanism. Using a numerical example, we show the advantages of using information from smart sensors to help multi-Kanban mechanism in increasing production efficiency.
The collection of the high value transport packaging is often ignored. There are times when these assets are lost in the logistics networks. As a result, companies are unable to gain full benefit from multiple uses of these packaging. One promising and emerging technology that has the potential of achieving an effective implementation and control of packaging is radio-frequency identification (RFID), which employs wireless communication between a tag and a reader to provide hands-off monitoring of transport packaging. To this end, RFID offers benefits not only to a firm in cost saving from maintaining minimal stock of packaging, but also to destination customers in terms of accurate and timely delivery. This paper addresses the benefits of using RFID and performs a cost-benefit analysis to support the RFID argument.
We study the disassembly process of a moderately complex consumer product (cell phone) consisting of 25 components. The investigation involves a heuristically solved disassembly line balancing problem and the determination of an appropriate disassembly sequence influenced by sequence-dependent costs, which is based on a novel combination of exact and heuristic algorithms. The issue of generating various suboptimum solutions to cope with multiple optimization criteria is also discussed.
Not many quantitative models in the literature deal with the strategic planning of a closed loop supply chain. They either deal with the forward supply chain or the reverse supply chain, but not both together. While the models dealing with forward supply chain do not address the issue of environmental consciousness, the models dealing with reverse supply chain assume that each incoming product is economical to re-process and each available recovery facility is efficient enough to re-process the incoming used products. With an objective to address these drawbacks, in this paper, we propose a multi-phase multi-criteria approach for the efficient design of a closed-loop supply chain. In the first stage, we formulate a linear integer programming problem to select economical products to process in the supply chain. In the second stage, we employ Analytic Network Process coupled with the Extent Analysis method to derive impacts for the criteria in selecting potential production facilities. In the final stage, we formulate a physical programming problem to obtain the transportation of right mix and quantities of goods across the supply chain. The methodology is illustrated with numerical examples.
Reverse logistics is the movement of used products from consumers towards reprocessors (re-manufacturers/ recyclers) within a supply chain. Possible drivers for companies interested in used products are recoverable value through re-processing, environmental regulations, product stewardship, and asset protection. Any reverse supply chain consists of at least three parties: collection centers where consumers return used products, recovery facilities where re-processing is performed, and demand centers where customers buy re-processed goods viz., output of recovery facilities. Strategic planning (also called designing) is long-range planning and is typically performed every few years when a supply chain needs to expand its capabilities. This paper suggests approaches for some crucial issues in strategic planning of a reverse supply chain.
In this paper, we concentrate on the disassembly-to-order (DTO) system, where end-of-life (EOL) products are taken back from last users to be disassembled to fulfill the demands for components and materials. The objective is to determine the number of EOL products that would be needed to maximize the profit and minimize the costs of the system. The conditions of EOL products are not always certain, which makes the problem difficult. We use a heuristic approach which transforms the stochastic disassembly yields into their deterministic equivalents and use a multi-criteria decision-making technique to solve the problem. In addition, we take the products’ ages (and thus their deterioration) into account to determine their yield rates (e.g., older products tend to have lower yield rates for usable components) and generate the DTO plans for multiple periods. A numerical example is considered to illustrate the implementation of the approach.
The automobile manufacturing industry plays a very important role in a country’s economy. The importance of automobile manufacturing industry lies in its sheer size and complexity in terms of the direct and indirect influence it commands across many other industries. While millions of people are employed in the automobile manufacturing industry, it is estimated that more than two and half times that number are employed in the auxiliary companies that supply parts to the automobile manufacturing companies. The auxiliary companies represent a group of businesses of various sizes, types, and geographical locations, producing a vast variety of products ranging from the very simple to the extremely intricate. In this study, the current environmental practices of management in the core Spanish auxiliary companies that do business with the automobile manufacturing industry (and thus form a large part of the automobile manufacturing industry’s supply chain) are investigated. We show that while automobile manufacturing companies are under scrutiny to become more and more environmentally friendly, not only at their manufacturing stage but also at their products’ useful and EOL stages, there appears to be no such burden on the auxiliary companies. Our conclusion is based on an elaborate survey conducted during the fall of 2004 of Spanish auxiliary companies with questions about the characteristics, environmental practices and reverse logistics related activities carried out by the companies.
The use of household appliances continues to rise every year. A significant number of End-Of-Life (EOL) appliances are generated because of the introduction of newer models that are more attractive, efficient and affordable. Others are, of course, generated when they become non-functional. Many regulations encourage recycling of EOL appliances to reduce the amount of waste sent to landfills. In addition, EOL appliances offer the appliance manufacturing and remanufacturing industries a source of less expensive raw materials and components. For this reason product recovery has become a subject of interest during the past decade. In this paper, we study the disassembly line for appliance disassembly. We discuss and incorporate some of the complications that are inherent in disassembly line including product arrival, demand arrival, inventory fluctuation and production control mechanisms. We show how to overcome such complications by implementing a multi-kanban system in the appliance disassembly line setting. The multi-kanban system (MKS) relies on dynamic routing of kanbans according to the state of the system. We investigate the multikanban mechanism using simulation and explore the effect of product mix on performance of the traditional push system (TPS) and MKS in terms of controlling the system’s inventory while attempting to achieve a decent customer service level.
The growing desire of consumers to acquire the latest technology (both at home and in the workplace), along with the rapid technological development of new products, has led to a new environmental problem: waste. The only way to tackle this problem is design and implementation of reverse supply chains. Implementation of an efficient reverse supply chain requires coordination among a number of parties, such as the collector, the dismantler, the shredder, and the recycler. In this paper, we identify four different scenarios of homogeneous and heterogeneous products, and formulate some potential interactions between the collector and the dismantler, for each of those scenarios.
In this paper, we employ fuzzy AHP methodology for selecting potential recovery facilities in a closed-loop supply chain. This methodology utilizes triangular fuzzy numbers for pair-wise comparisons and the extent analysis method for the synthetic extent value of the fuzzy pair-wise comparisons and principle of comparison of fuzzy numbers to derive the weight vectors to address the criticism traditional AHP often faces due to its unbalanced scale of judgments and inability to handle inherent uncertainty in carrying out pair-wise comparisons. A numerical example is considered to illustrate the methodology.
This paper deals with disassembly sequencing problems subjected to sequence dependent disassembly costs. We present a heuristic and an iterative method based on partial branch and bound concept to solve such problems. Since heuristic methods intrinsically generate suboptimum solutions, we compared the heuristically obtained solutions with the exact solutions to see if they are reasonably good or not. This process, however, is limited to small or perhaps medium sized problems only as the required CPU time for exact methods tends to increase exponentially with the problem size. For the problems tested, we observed that the methods described in this paper generate surprisingly good results using almost negligible amount of CPU time.
Disassembly takes place in remanufacturing, recycling and disposal, with a line being the best choice for automation. The disassembly line balancing problem seeks a sequence which: is feasible, minimizes workstations, and ensures similar idle times, as well as other end-of-life specific concerns. Finding the optimal balance is computationally intensive due to exponential growth. Combinatorial optimization methods hold promise for providing solutions to the disassembly line balancing problem, which is proven here to belong to the class of unary NP-complete problems. Probabilistic (ant colony optimization) and uninformed (H-K) search methods are presented and compared. Numerical results are obtained using a recent case study to illustrate the search implementations and compare their performance. Conclusions drawn include the consistent generation of near-optimal solutions, the ability to preserve precedence, the speed of the techniques, and their practicality due to ease of implementation.
Product and material recovery relies on the disassembly process to separate target components or materials from the end-of-life (EOL) products. Disassembly line is especially effective when products in large quantity are disassembled. Unlike an assembly line, a disassembly line is more complex and is subjected to numerous uncertainties including stochastic and multi-level arrivals of component demands, stochastic arrival times for EOL products, and process interruption due to equipment failure. These factors seriously impair the control mechanism in the disassembly line. A common production control mechanism is the traditional push system (TPS). TPS responds to the aforementioned complications by carrying substantial amounts of inventories. An alternative control mechanism is a newly developed multi-kanban pull system (MKS) that relies on dynamic routing of kanbans, which tends to minimize the system's inventories while maintaining demand serviceability. In this paper we explore the impact of sudden breakdown of server on the performance of a disassembly line. We compare the overall performances of the TPS and MKS by considering two scenarios. We present the solution procedure and results for these cases.
In this paper, we compare the impact of different disassembly line balancing (DLB) algorithms on the performance of our recently introduced Dynamic Kanban System for Disassembly Line (DKSDL) to accommodate the vagaries of uncertainties associated with disassembly and remanufacturing processing. We consider a case study to illustrate the impact of various DLB algorithms on the DKSDL. The approach to the solution, scenario settings, results and the discussions of the results are included.
This paper deals with a cost management problem of a remanufacturing system with stochastic demand. We model the system with consideration for two types of inventories. One is the actual product inventory in the factory. The other is the virtual inventory that is being used by the customer. For this virtual inventory, it should be required to consider an operational cost that we need in order to observe and check the quantity of the inventory. We call this the virtual inventory cost and model the system by including it. We define the state of the remanufacturing system by the two inventory levels. It is assumed that the cost function is composed of various cost factors such as holding, backlog and manufacturing costs. We obtain the optimal policy that minimizes the expected average cost per period. Numerical results reveal the effects of the factors on the optimal policy.
The main objective of a product recovery facility (PRF) is to disassemble end-of-life (EOL) products and sell the reclaimed components for reuse and recovered materials in second-hand markets. Variability in the inflow of EOL products and fluctuation in demand for reusable components contribute to the volatility in inventory levels. To stay profitable the PRFs ought to manage their inventory by regulating the price appropriately to minimize holding costs. This work presents two deterministic pricing models for a PRF bounded by environmental regulations. In the first model, the demand is price dependent and in the second, the demand is both price and time dependent. The models are valid for single component with no inventory replenishment sale during the selling horizon . Numerical examples are presented to illustrate the models.
Analytic Hierarchy Process (AHP) has been employed by researchers for solving multi-criteria analysis problems. However, AHP is often criticized for its unbalanced scale of judgments and failure to precisely handle the inherent uncertainty and vagueness in carrying out the pair-wise comparisons. With an objective to address these drawbacks, in this paper, we employ a fuzzy approach in selecting potential recovery facilities in the strategic planning of a reverse supply chain network that addresses the decision maker’s level of confidence in the fuzzy assessments and his/her attitude towards risk. A numerical example is considered to illustrate the methodology.
The disassembly line is the best choice for automated disassembly of returned products, however finding the optimal balance is computationally intensive. A greedy algorithm is presented for obtaining solutions to the disassembly line balancing problem, addressing hazardous and high-demand components. A two-optimal algorithm is developed to balance the part removal sequence. In addition, a hill-climbing heuristic is proposed to more rapidly balance the sequence. Examples are considered to illustrate the methodology. The conclusions drawn include the consistent generation of optimal or near-optimal solutions, the ability to preserve precedence, the speed of the algorithms and their practicality due to ease of implementation.
Gupta and Taleb defined and characterized the disassembly scheduling problem dealing with a single product type with assembly structure. Since the problem can be regarded as a reversed form of Material Requirement Planning (MRP), they suggested a Reverse MRP algorithm. However, they did not take into account one of the most important factors of scheduling, that is, how to carry out lot-sizing in Reverse MRP. The objective of this work is to complete that previous definition of the Reverse MRP algorithm with an easy eight-step procedure to determine how applying lot-sizing in Reverse MRP, that is to say, the timing and magnitude of orders to satisfy demand of leaf items (recycling or reuse components) considering ordering and holding costs thus joining one or more of the net requirements.
This paper presents a multi-criteria optimization model of a disassembly-to-order (DTO) system under uncertainty. The goal of the proposed model is to determine the best combination of the number of each product type to be taken back from the last user and/or collectors. The EOL products are then disassembled for the retrieval of reusable components and materials and resold in order to meet a certain level of demand under a variety of physical, financial and environmental constraints. The surplus components are recycled, stored for usage in subsequent periods or properly disposed. The problem is modeled as a multi-criteria decision-making problem under uncertainty, where the aspiration levels for various goals are more likely to be in the “approximately more (less) than” and/or “more (less) is better” form. We employ fuzzy goal programming technique to solve the problem. When solved, the model provides the number of EOL products to be taken back as well as the number of items reused, recycled, stored and disposed. The values of a host of other performance measures are also obtained, including total profit, materials and items sales revenues, take back cost, transportation costs as well as costs of preparation of EOL products, destructive disassembly, non-destructive disassembly, recycling, storage and disposal. A case example is presented to illustrate the model’s implementation.
In this paper, we investigate effect on the performance of two production control mechanisms, viz. push and pull control mechanism in a disassembly line where interruptions caused by preventive maintenance are expected. In the push system, blocking and starvation are compensated by the large amount of inventory stored in the buffer. Thus, the system must disassemble components prior to the arrival of actual demand. Alternatively, the multi-kanban pull system offsets the blocking and starvation by manipulating kanban routing. Hence, it directs kanban to the proper workstations based on system’s current status. Comparing with the push system, the system tends to keeps smaller amount of inventory while attempting to service the actual demand. Through numerical examples, we present advantages and disadvantages of both systems.
In this paper, a single phase Linear Physical Programming (LPP) model is formulated that explores the strategic and tactical planning stages of a Closed-Loop Supply Chain Network (CLSC). The model when solved identifies simultaneously the most economical used-product to re-process in the closed-loop supply chain, the efficient production facilities and the right mix and quantity of goods to be transported across the supply chain. A numerical example is considered to illustrate the methodology.
A three-phase mathematical programming approach for the efficient design of a closedloop supply chain network (CLSC) is proposed. Phase 1 deal with the selection of the most economical used-product to re-process in a CLSC, for which a linear physical programming (LPP) model is formulated. In phase 2, a fuzzy multi-criteria analysis approach is developed for selecting potential production facilities. In phase 3, a linear integer programming model is formulated to obtain the transportation of right mix and quantities of goods across the supply chain.
Apart from the efficient design, the success of a closed-loop supply chain network depends on its marketing strategy as well. Hence, it is important that the planned marketing strategy be evaluated with respect to the drivers of public participation in the network. To this end, we identify the important drivers of public participation and propose a fuzzy analytic network process (ANP) based methodology to evaluate the marketing strategy of a closed-loop supply chain with respect to those drivers. A numerical example is considered to illustrate the methodology.
Strategic planning of a supply chain primarily deals with the design (what products should be processed/produced in what facilities etc) of the supply chain that is typically a long-range planning. Tactical planning involves the optimization of flow of goods and services across the supply chain and is typically a medium-range planning. In this paper, we present a single-phase unified approach, employing goal programming, for these two stages of planning of a Closed- Loop Supply Chain Network (CLSC). When solved, the model identifies simultaneously the most economical used-product to re-process in the supply chain, the efficient production facilities and the right mix and quantity of goods to be transported across the supply chain. A numerical example is considered to illustrate the methodology.
In this paper, we consider the disassembly-to-order (DTO) problem, where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The main objective is to determine the optimal number of take-back EOL products for the DTO system that satisfy the desirable criteria of the system. Since the objectives of the DTO system always have different priorities and uncertain aspiration levels, weight-priority assignment and fuzzy sets theory have been applied to a multi-criteria decision-making approach to solve the problem. In addition, we consider product’s deterioration that affects the product’s yield rates (e.g., older products tend to have lower yield rates for usable components) and use heuristic procedure to transform the stochastic disassembly yields into their deterministic equivalents. The DTO plan is generated for multiple periods. A numerical example is also considered.
In this paper, we focus on solving the disassembly-to-order (DTO) system, where end-of-life (EOL) products are taken back from last users to be disassembled to fulfill the demands of components and materials. The objective is to determine the optimal number of EOL products for the DTO system that satisfy the desirable criteria of the system. Frequently, decision makers (DM) find it difficult to specify the aspiration levels and weights relating to the priority of each objective of the DTO system. However, linear physical programming (LPP), which has key features to entirely remove the DM from the process of choosing weights and to handle the vagueness of aspiration levels, can be very suitable in such a situation. In this paper we generate the DTO plan for multiple periods using LPP. A numerical example is considered to illustrate the methodology.
Product disassembly takes place in remanufacturing, recycling, and disposal. The disassembly line is the best choice for automated disassembly, so it is essential that it be designed and balanced to work efficiently. A disassembly line faces many unique circumstances including: inventory problems because of the disparity between demands for parts and their yield, uncertainty in the returned product’s structure and quality, and polluting or hazardous parts. The multi-criteria Disassembly Line Balancing Problem seeks to: provide a feasible disassembly sequence, minimize the number of workstations, minimize idle time, and balance the line (ensure similar idle times at each workstation) as well as addressing other, disassembly-specific concerns. However, finding the optimal balance is computationally intensive – due to factorial growth and exhaustive search quickly becoming prohibitively large – necessitating the use of heuristics. This paper demonstrates the mathematical foundations for the multiple objective Disassembly Line Balancing Problem using original mathematical formulae including those for determining upper and lower theoretical bounds. These formulae are essential in enabling a thorough efficacy analysis of a given part removal solution sequence. Because heuristics typically generate solutions that are less than optimal, metrics are needed to compare heuristics to each other or to a known best and worst case. In addition, one of the concerns when using heuristics is the idea that very little has been rigorously established in reference to their performance. Therefore, developing ways of explaining and predicting their performance is considered to be one of the most important challenges currently facing the fields of optimization and algorithms. In this paper, mathematical tools for quantitative measurement and graphical tools for qualitative analysis are developed and reviewed, focusing on analytical methodologies used in evaluating heuristic searches.
Disassembly line is crucial in recovering products in large quantities. It has recently gained a lot of attention due to its role in efficiently recovering valuable materials, components, and subassemblies. Disassembly involves many challenges that further complicate the process. Recent research reported in the literature has focused on balancing disassembly lines. That means the time required for the work content at each station is balanced. However, the inventory generated at various stations may not be balanced. Additional inventory problems arise because of the disparity between the demand of certain components and the actual yield from the products. In this paper, we discuss these and other inventories issues to provide a better understanding of complications resulting from this and provide an approach to overcome them. A numerical example is considered to illustrate the approach.
Product recovery facilities that collect end-of-life (EOL) products to recover reusable components often claim that processing EOL products is an economically risky proposition. The major factor that contributes to the low pro¯t margin is the °uctuating inventory levels caused by the cumulative e®ect of inconsistent in°ows of EOL products and varying demand patterns. By adopting a suitable pricing strategy, one can control the variability in the in°ow without compromising on the revenue. This research presents pricing strategies for three scenarios that are formulated as optimal control problems with inventory constraints. In the ¯rst scenario the PRF seeks to keep its disposable inventory below the stipulated disposal limit for the entire span of the selling horizon. The second scenario captures the situation where the PRF targets to achieve a zero inventory at the end of the selling horizon. In the third scenario the PRF exercises control over the production rate of reusable items and desires to determine the optimal production quantity and optimal price during the selling horizon. Examples are presented to illustrate e®ectiveness of the pricing model for each scenario.
This paper describes the implementation of JIT in a small manufacturing company and the benefits that resulted for the company's operations. Preliminary analysis identified various problems in the existing manufacturing operations. The pre-implementation and post-implementation conditions of the company are detailed. The achievements of the JIT implementation included a reduction in material traversing, reduced lead times and inventories leading to an overall reduction in the cost of manufacturing. A smooth flow of material from the raw material stage through the finished product stage was established. Three separate product lines were combined into a flexible manufacturing assembly line. With the adoption of a holistic approach to JIT implementation, it was found that even a small company can make significant strides towards world class manufacturing status. The experience gained by the company can encourage and benefit other small companies to embrace the JIT approach.
Variability in the inflow of end-of-life (EOL) products and fluctuating inventory levels often make the processing of EOL products an economically risky operation for product recovery facilities (PRFs). Choosing an appropriate pricing policy can enhance the performance of PRFs by methodically clearing their inventory and increasing profits. This work presents two pricing models to counter the prospect of product obsolescence that can happen either gradually or suddenly. Product obsolescence can cause demand drop and inventory pile up, both of which could dent the revenues of PRFs. In the first model, gradual obsolescence and environmental regulations that limit the disposal quantity in landfills are considered. In the second model, the case of sudden obsolescence is addressed. Examples are presented to illustrate the pricing strategies for each model.
Economic incentives, government regulations and customer perspective on environmental consciousness (EC) are driving more and more companies into the product recovery business, which forms a reverse supply chain. The combination of traditional/forward supply chain and reverse supply chain is called a closed-loop supply chain (CLSC). A Supply Chain involves three stages of planning, viz., Strategic, Tactical and Operational. Strategic planning primarily deals with the design (what products should be processed/produced in what facilities etc) of the supply chain that is typically a long-range planning performed every few years when a supply chain needs to expand its capabilities. Tactical planning involves the optimization of flow of goods and services across the supply chain and is typically a medium-range planning performed on a monthly basis. Finally, Operational planning is a short-range planning that deals with the day-to-day production planning and inventory issues on the factory floor. In this paper, we formulate a single-phase linear physical programming model in designing a CLSC. This model when solved addresses simultaneously the critical issues in the strategic and tactical planning of a CLSC. The criteria considered in the problem formulated include, the used-product collection cost at the collection centers, disassembly and remanufacturing costs at the production facility, new products production cost at the facility, transportation costs across the supply chain, inventory carrying costs, disposal cost of broken/unfit used-products and revenues from the sale of remanufactured products, new products and recycling of used-products that are not fit for remanufacturing but have some residual material value.
The first crucial step of product recovery is disassembly. A disassembly sequence plan is a sequence of disassembly tasks, which begins with a product to be disassembled and terminates in a state where all of the parts of interest are separated. The problem of finding the optimal disassembly sequence plan is NP-complete problem and therefore complex and challenging to solve. In this paper, we seek a Disassembly Sequence Plan that addresses two criteria in order: first, we look for a sequence, the cost of which is close to our cost aspiration; secondly, we look for a sequence that prioritizes some selected parts to be disassembled as early as possible. We propose a Greedy Randomized Adaptive Search procedure (GRASP) based heuristic methodology specifically developed to solve such bi-criteria type of disassembly problem. Some experiments were designed testing the good performance of the approach.
In this paper, we consider the disassembly-to-order (DTO) problem, where a variety of returned products are disassembled to fulfill the demand for specified numbers of components and materials. The objective is to determine the optimal numbers of returned products to disassemble so as to maximize profit and minimize costs. We model the DTO problem using a multi-criteria decision-making approach. Since the conditions of returned products are unknown, the yields from disassembly are considered to be stochastic. To solve the stochastic problem, we use one of the two heuristic approaches (viz., one-to-one approach or one-to-many approach) that convert the problem into a deterministic equivalent. We compare the performance of the two heuristic approaches using a case example.
Disassembly takes place in remanufacturing, recycling, and disposal, with a line being the best choice for automation. The disassembly line balancing problem seeks a sequence, which is feasible, minimizes the number of workstations, and ensures similar idle times as well as other end-of-life specific concerns. Finding the optimal balance is computationally intensive due to exponential growth. Combinatorial optimization methods hold promise for providing solutions to the problem, which is proven here to be NP-hard. Stochastic (genetic algorithm) and deterministic (greedy/hillclimbing hybrid heuristic) methods are presented and compared. Numerical results are obtained using a recent electronic product case study.
Product recovery facilities (PRFs) usually report low profit margins because the quantity and timing of end-of-life product returns are unpredictable and the demand for reusable items fluctuates without a fixed pattern. In a situation like this selling items in inventory at suitable prices can minimize inventory variations and increase profits. This work presents an optimal pricing policy for reusable items considering the local environmental legislations that restrict the PRF’s disposal quantity. The effect of initial inventory, disposable inventory, disposal limit, production cost, disposal cost, and holding cost on prices, items sold, and profit is also studied.
Product disassembly takes place in remanufacturing, recycling, and disposal. The disassembly line is the best choice for automated disassembly, so it is essential that it be designed and balanced to work efficiently. The multiobjective disassembly line balancing problem seeks to find a disassembly sequence which provides a feasible disassembly sequence, minimizes the number of workstations, minimizes idle time, balances the line (ensures similar idle times at each workstation), as well as addressing other disassembly-specific concerns. However, finding the optimal balance is computationally intensive due to exponential growth, with exhaustive search quickly becoming prohibitively large. In this paper, an ant colony optimization metaheuristic is presented for obtaining optimal or near-optimal solutions to the disassembly line balancing problem. Examples are considered to illustrate implementation of the methodology. Conclusions drawn include the consistent generation of near-optimal solutions, the ability to preserve precedence, the superior speed of the metaheuristic, and its practicality due to its ease of implementation.
At the end-of-life (EOL) of a product, there are several options available for its processing including reuse, remanufacturing, recycling and disposing. In almost all cases, a certain level of disassembly may be necessary. Thus, finding an optimal (or near optimal) disassembly sequence is crucial to increasing the efficiency of the process. Disassembly operations are labor intensive, can be costly, have unique characteristics and cannot be considered as the reverse of assembly operations. Since the complexity of determining the best disassembly sequence increases with increase in the number of parts of the product, it is extremely crucial that an efficient methodology for disassembly sequencing be developed. In this paper, we present a genetic algorithm for disassembly sequencing of EOL products. A case example is considered to demonstrate the functionality of the algorithm.