多产品串并联生产系统动态看板的设计和计算

针对机电产品企业普遍采用的多品种、串并联生产系统,提出了批量轮番生产方式下采用动态看板作为控制手段时看板内容的设置和看板数量的计算方法,使得看板的内容和数量能随“需求”而动态变化,实现了“多品种均衡”生产和现场在制品数量的有效控制。     

Determining inventory levels in a CONWIP controlled job shop

We extend the concept of CONWIP control to a job shop setting, in which multiple products with distinct routings compete for the same set of resources. The problem is to determine the fixed overall WIP level and its allocation to product types (WIP mix) to meet a uniformly high customer service requirement for each product type. We formulate an optimization problem for an open queuing network model in which customer orders pull completed products from the system. Then, assuming heavy demand, we derive a throughput target for each product type in a closed queuing network and provide a simple heuristic to find a minimum total WJP and WIP mix that will achieve an operating throughput close to this target. In numerical examples, the WIP mix suggested by this approach achieves the customer service requirement with a relatively low total WIP     

DETERMINING THE NUMBER OF KANBANS IN MULTI-ITEM JUST-IN-TIME SYSTEMS

Kanbans (cards) are one means of controlling Just-in-time production systems. The kanbans serve as production authorization orders and essentially become the information system. The presence of a production ordering kanban for a specific part type constitutes an instruction to the workcenter to produce a “container” of that part. In this paper we assume workcenters produce multiple part types and address the issue of the number of kanbans needed for each part type. Our objective is to minimize the sum of inventory holding and backorder cost. A stochastic model is formulated. Steady-state results are derived for the cases of a few and many part types. Modifications for the case of expediting of backorders are also provided. Simulation results demonstrate the accuracy of the model for a large number of kanbans, but an investigation of a finite population queueing model for small numbers of kanbans is suggested as a worthwhile extension.     

Flexible kanbans to enhance volume flexibility in a JIT environment: a simulation based comparison via ANNs

Kanbans play an important role in the information and material flows in a JIT production system. The traditional kanban system with a fixed number of cards does not work satisfactorily in an unstable environment. In the flexible kanban-type pull control mechanism the number of kanbans is allowed to change with respect to the inventory and backorder level. Based on the need for the flexible kanban, a method was proposed by (Husseini, S.M.M., O’Brien, C., and Hosseini, S.T., 2006. A method to enhance volume flexibility in JIT production control. International Journal of Production Economics, 104 (2), 653–665), using an integer linear programming technique, to flexibly determine the number of kanbans for each stage of a JIT production system, minimising total inventory cost for a given planning horizon. Here, the effectiveness of the method proposed by Husseini et al. is examined by a case study and compared with the results for the conventional method of fixed kanban determination. This is also confirmed by a simulation study using artificial neural networks (ANNs). The main aim of this paper is to show the cost advantage for Husseini et al.'s method over the conventional method in fluctuating demand situations, and especially to prove that simulation via ANNs ensures a simplified representation for this method and is time saving.     

A decomposition method for multi-product kanban systems with setup times and lost sales

We consider kanban controlled production systems with three or more different products processed on a single manufacturing facility. Customers for a product arrive according to a Poisson process. If a customer's demand cannot be met from stock, the customer leaves and satisfies his demand elsewhere (lost sales). Between the production of different products setup changes must be performed that take a significant time. Setup times and processing times are product-specific and follow exponential distributions. A production run continues until the target inventory level given by the number of kanbans for the product has been reached (exhaustive processing). Then the manufacturing facility is set up for the next product according to a fixed setup cycle. The exact model is mathematically intractable even for smaller systems. Therefore, we propose a decomposition-based approximation method for estimating steady-state performance measures     

Performance evaluation of two-stage multi-product kanban systems

We propose an approximate evaluation procedure for a kanban-controlled production system with two stages and multiple products. In the first stage, single-product manufacturing facilities produce items that are the input material for a single multi-product manufacturing facility in the second stage. Each manufacturing facility is controlled by a distinct kanban loop with a fixed number of kanbans. Processing and setup times are exponentially distributed, demand arrivals at the output store of the second stage are Poisson and independent for each product. If a customer's demand cannot be met from stock, the customer either waits or leaves the system, depending on the admissible number of backorders and the current number of waiting customers (partial backordering). We describe a new decomposition-based approximation method for the evaluation of such systems in steady state. We focus on the performance measures average fill rate, average fraction of served demand, and average inventory level. We report the results of several numerical tests. The results indicate that the approximation is sufficiently accurate for a large variety of systems. We also illustrate the effects of increasing the maximum number of backorders on the performance of the system.     

Determining the number of kanbans: a step toward non-stock-production

Just-in-time (JIT) production is a philosophy that calls for reducing work-in-process (WIP) inventory to aid process improvement and reduce process variability. In some cases, JIT production has been misinterpreted as a method that would lead to zero or minimal WIP with a lot size of one. There are no models or theories to achieve the JIT goal, i.e. non-stock-production (NSP), and, in particular, to help determine when and where to maintain this minimal inventory. A kanban system acts as the nerve of a JIT production system whose functions are to direct materials just-in-time to workstations in stages of manufacturing, and to pass information as to what and how much to produce. Indeed, the number of kanbans between two adjacent workstations decides the inventory level of that pair of workstations. With the objective of minimizing WIP inventory level, one model dealing with three cases of production configuration is developed for deciding the optimum number of kanbans. The model is then solved using a Markov process approach which considers the demand of finished products as the departure rate and the production rates of stations as arrival rate. In this paper, the model and solution procedure are illustrated with a numerical example.     

THE KANBAN ASSIGNMENT PROBLEM IN SERIAL JUST-IN-TIME PRODUCTION SYSTEMS

This paper examines multi-stage serial just-in-time production systems that use kanbans as a means of communicating the production needs of each stage of operation. The kanban-controlled system is a linked-demand system. A recursive function is presented which computes the number of kanbans required for all stages of a serial production system. This function is used to compute a lower bound on the expected fill rate for an arbitrary kanban assignment.     

Algorithms for the design of a multi-stage adaptive kanban system

The traditional kanban system with a fixed number of cards does not work satisfactorily in an unstable environment. With the adaptive kanban-type pull control mechanism, the number of kanbans is allowed to change with respect to the inventory and backorder level. It is required to set the threshold values at which cards are added or deleted, which is part of the design. Previous studies used local search and meta-heuristic methods to design an adaptive kanban system for a single stage. In a multi-stage system the cards are circulated within the stage and their presence at designated positions signals to the neighbouring stages details concerning the inventory. In this work, a model of a multi-stage system adapted from a traditional and adaptive kanban system is developed. A genetic and simulated annealing algorithm based search is employed to set the parameters of the system. The results are compared with a traditional kanban system and signs of improvement are found. The numerical results also indicate that the use of a simulated annealing algorithm produces a better solution.     

Joint lot-sizing and scheduling for multi-stage multi-product flow shops

Here we consider a multi-product flow shop under capacity constraints and develop an efficient heuristic procedure to determine joint lot sizes, production sequence and production schedule. We treat setup costs as fixed in the short run and thus independent of the number of setups. Loss of production capacity due to setup times is explicitly accounted for and the transfer of portions of a production lot between stages is permitted. The procedure is based on identifying the bottleneck workcentre and synchronizing production schedules at all other workcentres with the bottleneck workcentre such that product throughput requirements are met with minimal inventory costs. A compact procedure for constructing the Gantt charts is also presented. Conversion of the lot si2es and transfer batches into an information control system with kanbans is also presented.     

Evaluation of performance measures for multi-part,single-product kanban controlled assembly systems with stochastic acquisition and production lead times

We consider the following kanban-controlled, multi-stage production assembly system. A number of raw parts are acquired from various suppliers and assembled into a single product. The raw-part acquisition lead times, the production lead time and demand arrival are all random variables. The raw-part acquisition order is made when the inventory level of the common part buffer, consisting of a number of sets of raw parts where a set of raw parts forms a single product, depletes to a reorder point. A production stage consists of the input queue, the output buffer, and the kanban board. The finished product can be backordered with a given allowable quantity. The problem is to evaluate the various system performance measures for a given set of design parameters: the raw-part batch order size, the common buffer size, the reorder point, the number of kanbans circulating in each production stage. A system is decomposed into a number of semi-autonomous Markov processes. A mathematical model is formulated and an iterative algorithm is proposed to evaluate system performance measures. Extensive numerical experiments with simulation analyses show that the computational time is very short and the model proposed is reasonably accurate. Thus, resorting to a genetic algorithm we can find an optimal set of system design parameters.     

Optimal method for changing the number of kanbans in the e -Kanban system and its applications

The Toyota Motor Corporation has recently developed a new kanban system called ‘e-Kanban’, which is a parts ordering information system that operates within the communications network established between Toyota and its suppliers. One of the goals for developing the e-Kanban system was to introduce an efficient means for properly changing the number of kanbans once the required number has been calculated, as changing the number of kanbans in the kanban system directly affects the order quantity. Consequently, it is very important to monitor carefully and control changes in the number of kanbans. The paper investigates and proposes an effective method for changing the number of kanbans using the e-Kanban system. Applying this method to three problems, it is shown that the e-Kanban system can be implemented more efficiently and effectively than the original kanban system.     

A simulative approach for evaluating alternative feeding scenarios in a kanban system

In accordance with the lean production philosophy, an assembly line may be supplied by means of a kanban system, which regulates and simplifies the flow of materials between the lines and the warehouses. This paper focuses on evaluation of feeding policies that differ from each other in term of the number of kanbans managed per feeding tour. A pure cost-based approach is thus proposed, which considers both inline inventories along with handling costs proportionate to the number of operators involved in the parts-feeding process. A multi-scenario simulative approach is applied in order to establish the number of operators required to avoid inline shortages. The scenario minimising total cost is then selected. The innovation introduced is a model for describing kanban arrivals and their requests for feeding, improving the potential of the simulation to describe real-life environments. Lastly, a case study from the automotive industry is presented in order to highlight the applicability of the proposed approach as well and the effects of alternative feeding policies on the total cost incurred.     

Analysis of production control systems kanban and CONWIP

In this paper we describe and classify different pull production systems. The production control systems kanban and CONWIP are then analysed with respect to production rate and average W1P. We examine single product flow lines with exponential service time distributions and unlimited demand at the final buffer (saturated lines). We show that the distribution of cards (kanbans) has a significant effect on the performance of kanban systems. Different types of kanban control mechanisms show equivalent performance data, if the distribution pattern is adapted accordingly. Our research shows that the kanban system is more flexible with respect to a given objective than the CONWIP system, because in addition to the total card number the card distribution is a control parameter. Moreover for a given production rate the average WIP is lower in a kanban system than in a CONWIP system. We identify the average WIP in the interstage buffers as important parameter and describe the ‘WIP bowl phenomenon’ as result of optimum card distributions. Concluding remarks and directions for future research conclude the paper.     

Effects of kanban withdrawal policies and other factors on the performance of JIT systems—a simulation study

This paper presents the results of a simulation study of a just–in–time (JIT) production control system and its performance under different operational conditions. In particular, the effects of two different kanban withdrawal policies on such performance measures as throughput rate, station utilizations and total work in process levels are investigated. Also, simulation experiments are carried out to determine the effects of processing time variability, number of different types of kanbans allowed at each station and production line length on the mentioned performance measures of JIT production control method. The simulation experiments were carried out for production lines in which processing times of stations were assumed to follow gamma and Erlang distributions.     

Backorder minimization in multiproduct assemble-to-order systems

We consider a multiproduct assemble-to-order system. Components are built to stock with inventory controlled by base-stock rules, but the final products are assembled to order. Customer orders of each product follow a batch Poisson process. The leadtimes for replenishing component inventory are stochastic. We study the optimal allocation of a given budget among component inventories so as to minimize a weighted average of backorders over product types. We derive easy-to-compute bounds and approximations for the expected number of backorders and use them to formulate surrogate optimization problems. Efficient algorithms are developed to solve these problems, and numerical examples illustrate the effectiveness of the bounds and approximations.     

Yield management with downward substitution and uncertainty demand in semiconductor manufacturing

Motivated by the high yield variability in the semiconductor industry where the quality of the end products is uncertain and is graded into one of several quality levels according to performance before being shipped. We consider a dynamic multi-period yield management problem of a two-stage make-to-stock system faced by a semiconductor manufacturing firm. In the first stage, the firm invests in raw material before any actual demand is known, and produces multiple types of products with random yield rates because of the presence of randomness in the process. In the second stage, products are classified into different classes by quality, and allocated in a number of sequential periods. Demand is also random and can be classified into multiple classes corresponding to product levels. Demands can be upgraded when one type of product has been depleted. This paper presents a multi-period, multi-product, downward substitution model to determine the optimal production input and allocation of the different products to satisfy demands. The production and allocation problem is modelled as a stochastic dynamic program in which the objective is to maximise the profit of the firm. We show that the simple one step upgrade substitution policy is optimal, and the objective function is concave in production input. An iterative algorithm is designed to find the optimal production input and numerical experiments are used to illustrate its effectiveness.     

Storage Decisions in Information Systems

The problem is to dynamically store different data records in different storage devices in each period so as to minimize the total expected discounted cost over a planning horizon. Each device has a fixed total capacity, each record has a given storage space requirement, while the number of requests for each record per period is changing stochastically through time. Given an allocation, the total cost per period consists of the storage cost (depending on the storage requirements and device), the access cost including update and retrieval costs (depending on the number of requests) and the transfer cost (depending upon the change of allocation from the previous period). A dynamic programming model is presented to yield optimal strategies. The special case of independent identically distributed demands is completely solved, using a generalized transportation algorithm while a heuristic procedure is indicated for the general problem using parametric analysis.     

Effects of group size and learning on manual assembly performance: an experimental study

In variable manual assembly production of highly customised products, effective allocation of workers to products is required. To support decision-making here, industrial managers should be aware of the performance effects of the number of workers and learning within individual products. Evidence on such fundamental effects requires laboratory studies with products similar to those in real assembly industries. Because of the lack of such studies, this paper studies experimentally the effects of group size (one to four workers) and learning (up to four repetitions per group) on the performance of product assembly. The product, built for the purpose of the present study, consists of representative elements from real products in the mechanical engineering industry. A total of 68 undergraduate students participated in the experiments. The results from the experiments are in line with the hypotheses that the mean assembly time decreases at a decelerating rate as a function of both group size and repetitions, and that productivity per worker decreases as a function of group size. The results are explained in more detail through the experiences of the participants. Managerial implications and aspects for future research are also discussed.     

Operator assignment with cell loading and product sequencing in labour-intensive assembly cells – a case study of a bicycle assembly company

This research deals with an operator assignment problem in which cell loading and product sequencing are taken into account in labour-intensive assembly cells. In each cell the number of assembly tasks is more than the number of operators, so multi-assembly tasks are assigned to each operator. Because the assembly procedure and time required for each assembly task are quite different for different products, some tasks will be transferred when the product changes. Reducing the number of task transfers can smooth the process of product change. A four-phase methodology is proposed to minimise the total manpower required and task transfers at the same time. The four phases are manpower configuration design, calculating the number of task transfers, manpower requirement minimization and cell loading and product sequencing optimization. A case study from a bicycle assembly company is introduced. For comparison, two methodologies are applied. The results show that the proposed four phase methodology can provided the solution with fewer task transfers based on the same total manpower requirement. However, when the number of products increases, the computation time of the proposed four-phase methodology increases rapidly.