تجزیه و تحلیل عملکرد و بهینه سازی سیستم های تولید هیبریدی تحت یک سیاست مرتب سازی دسته ای
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|3679||2013||9 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : International Journal of Production Economics, Available online 16 February 2013
We consider a stochastic two-stage hybrid manufacturing system for a single product where semi-finished goods are Made-To-Stock (MTS) and then differentiated when demand is realized through a Make-To-Order (MTO) stage. Inventory for semi-finished goods is held between the two stages. We introduce a batch ordering policy to permit economies of scale in ordering due to a cost associated with each order placed. We use the Matrix-analytic method to evaluate system performance under this ordering policy. Afterwards, we develop an optimization model to find the optimal intermediate buffer size and the optimal replenishment order quantity for the system. We show that a base stock policy is sub-optimal in the presence of a replenishment cost for semi-finished goods. The savings from adopting the batch ordering policy can be high while the response time for most customer orders is not affected.
Increased competition among manufacturing companies where flexibility, quality, cost and response time play a critical role is a major challenge in today's industry. Key to most companies is to increase product variety and yet maintain short delivery lead times. To achieve these goals, new manufacturing strategies have evolved. Hybrid manufacturing is one such trend where products are produced via a two-stage process: they are partially assembled using a Make-To-Stock (MTS) stage, and then customized based on customer requirements in a subsequent Make-To-Order (MTO) stage. Intermediate inventory of partially finished goods is stored in a buffer between the two stages. Hybrid MTO–MTS systems are one of the many interrelated and complementary strategies for designing and managing supply chains identified by Mikkola and Larsen (2004). These strategies are mass customization (see, e.g. Silveira et al., 2001), postponement and modularization. Hybrid MTS–MTO utilizes certain aspects of each of these three strategies to deliver customers with a customized product at minimal cost. MTS systems are usually suitable for high volume and low variety products, whereas MTO systems are suitable for low volume and high variety products. The hybrid manufacturing strategy, as suggested by Youssef et al. (2004), combines the benefits of both MTS and MTO. It reduces the order fulfillment delay relative to MTO because finished goods are made from partially completed items. At the same time, it can lower inventory costs since inventory is held only for semi-finished goods and raw components, rather than finished goods (Gupta and Benjaafar, 2004). The hybrid MTS–MTO strategy is widely used in the electronics industry and other similar markets where many product configurations can be produced from intermediate interchangeable modules (Gupta and Weerawat, 2006). A well-known and successful example of a company that has adopted this strategy is Dell Computer Corporation (Serwer, 2002). Research on MTS–MTO systems falls into four primary categories: The first is research on inventory management for the semi-finished goods between the two stages, e.g. Gupta and Benjaafar (2004). Second, research on the optimal point of differentiation, e.g. Gupta and Benjaafar (2004) and Jewkes and Alfa (2009). Third, research that explores the optimal configuration for semi-finished products from the point of view of commonality among different final products, e.g. Swaminathan and Tayur (1998). Finally, there is research that looks into coordination between the two stages, e.g. Gupta and Weerawat (2006). Our work falls in the first category. Most research on hybrid MTS–MTO manufacturing systems use a base stock, or one for one replacement policy for replenishing the intermediate buffer. That is, the MTO stage is triggered to replenish the semi-finished goods buffer by one unit each time a semi-finished item is removed from the intermediate inventory as a result of a customer order. This policy is widely adopted in the literature because of its ease of modeling, although such a policy is not necessarily optimal when there is an ordering cost associated with each replenishment. Veinott (1965) showed that when an ordering cost exists, a batch replenishment policy is optimal. In an MTS–MTO context, batch replenishment occurs when the MTS stage is signaled to restock the intermediate inventory only after a specified number of semi-finished items have been removed from the buffer. Compared to a base stock policy, batch replenishment may expose the system to longer customer lead times if the intermediate buffer is empty when a customer order is placed. This is because the customization process cannot begin until a unit of semi-finished goods is available. However, in the presence of replenishment costs, it is worth considering the impact of this tradeoff. The primary contribution of this work is to introduce a batch replenishment policy to hybrid MTS–MTO manufacturing systems. The second contribution is to provide an exact analysis for various performance measures that capture the tradeoffs between replenishment costs and customer lead times. Furthermore, we formulate an optimization problem to find the optimal values of the intermediate goods buffer and batch size decision variables. Our batch ordering policy is a more realistic generalization of the base stock policy and captures a common manufacturing problem faced by decision makers. While one might conclude that batching customer orders prior to issuing a replenishment order for intermediate goods would lead to increased customer lead times, our work shows that batching orders affect only a small percentage of customer orders. We are also able to quantify when the base stock policy produces sub-optimal solutions. To our knowledge, our paper is the first to examine the effects of batching replenishment orders when there are ordering/setup costs associated with each order placed for the semi-finished goods. The rest of the paper is organized as follows: In Section 2, we review some of the relevant literature. We describe the model in detail and develop a Markov chain representation in Section 3. In Section 4, we develop an optimization model to find the optimal intermediate buffer size and order batch size. Numerical analysis is reported in Section 5. Finally, we conclude our work with the main findings and point out future research opportunities in Section 6.
نتیجه گیری انگلیسی
Manufacturing systems, in the extreme, can be categorized as make-to-stock (MTS) or Make-To-Order (MTO). While MTS systems can respond to customer demand quickly, they are not suited for markets where customers require high variety in finished products for which inventory holding costs can be considerable. In contrast, MTO systems can provide highly customized products, but customers must wait for their orders to be filled. A hybrid MTS–MTO system can provide some of the positive characteristics of both approaches. Careful consideration of the semi-finished goods design for the MTS stage can combine with the MTO customization stage. This permits a wide variety of end products to be made from the standard semi-finished goods without the need to hold any finished goods inventory. Most research on the optimal design of MTS–MTO systems assume the use of a base-stock replenishment policy for the intermediate semi-finished goods buffer. While this is an optimal policy for systems with no replenishment costs, a batch ordering policy may outperform a base-stock policy in the presence of replenishment costs. This paper explored a possible way in which MTS–MTO systems could be adapted to take advantage of economies of scale in ordering semi-finished goods. The primary contribution of this work is to show the potential benefit of such batching and to demonstrate that there can be substantial savings to the manufacturer, but little cost to the consumer in terms of additional delays. One direction for future research is to consider adding a service level constraint to limit the extent to which batching increases customer delays. Other potential extensions are to model order-specific processing times at the MTO stage or where customer orders may consist of more than one item.