Inventory and Production Management in Supply Chains

Edward A. Silver, David F. Pyke, Douglas J. Thomas

• Éditeur: CRC Press
• Année: 2016
• ISBN:
  9781315356808
• Édition: 4
• Sujet: Betriebswirtschaft
• Sous-sujet: Operations

Authored by a team of experts, the new edition of this bestseller presents practical techniques for managing inventory and production throughout supply chains. It covers the current context of inventory and production management, replenishment systems for managing individual inventories within a firm, managing inventory in multiple locations and firms, and production management. The book presents sophisticated concepts and solutions with an eye towards today's economy of global demand, cost-saving, and rapid cycles. It explains how to decrease working capital and how to deal with coordinating chains across boundaries.

II

---

REPLENISHMENT SYSTEMS FOR MANAGING INDIVIDUAL ITEM INVENTORIES WITHIN A FIRM

---

In the first part of this book, we have laid groundwork by specifying the environment, goals, constraints, relevant costs, and forecast inputs of production/inventory decisions. Now, we turn our attention to the detailed logic for short-range decision making in situations of a strictly inventory (i.e., nonproduction) nature. We are concerned primarily with individual items or SKUs. However, as discussed in Chapter 2, aggregate considerations will often still play a crucial role by examining the consequences across a population of inventoried items of using different values of certain policy parameters. Management can select the aggregate operating point on an exchange curve, thus implying values of the parameters that are to be used in the item by item decision rules.

Within this part of the book, we deal with the so-called B items, which, as we have seen earlier, comprise the majority of items in a typical inventory situation. Unlike for C items, which will be handled in Section III, there are usually significant savings to be derived from the use of a reasonably sophisticated control system. On the other hand, the potential savings per item are not as high as those for individual A items. The combination of these two factors suggests the use of a hybrid system that combines the relative advantages of human and computer. The management-by-exception control system that we describe is reasonably sophisticated, and usually automated, but requires human intervention on a relatively infrequent basis. Such automated control systems, when applied in practice, have resulted in significant improvements in the control of B items.

In Chapter 4, we first answer the question of how much to replenish in the context of an approximately level demand pattern. Next, in Chapter 5, we consider the more general situation of a demand pattern, such as that seen with seasonal items, where the mean or forecast demand changes appreciably with time. This more general case considerably complicates the analysis. Therefore, the suggested decision system is based on a reasonable heuristic decision rule.

Chapter 6 is concerned, for the most part, with the situation where the average demand rate stays approximately level with time, but there is a random component present. That is, the mean or forecast demand stays constant or changes slowly with time, but there definitely are appreciable forecast errors. When demand is known only in a probabilistic sense, two additional questions must be answered: “How often should we review the stock status?” and “When should we place a replenishment order?”

All the material in Chapters 4 through 6 ignores the possible benefits of coordinating the replenishments of two or more items. We return to this important topic in Chapters 10 and 11 of Section IV.

Chapter 4

---

Order Quantities When Demand Is Approximately Level

---

In this chapter, we address the question of how large a replenishment quantity to use for a single item when conditions are rather stable (any changes in parameters occur slowly over time), and there is relatively little or no uncertainty concerning the level of demand. This simplified situation is a reasonable approximation of reality on certain occasions. More important, however, the results obtained turn out to be major components of the decision systems when parameters change with time, demand is probabilistic or multiple items are managed jointly. In addition, the goal is not to have a completely automated replenishment system, but to provide decision support to a planner. This is essential for several reasons, including (1) giving the planner the ability to incorporate factors not included in the underlying mathematical model, and (2) cultivating a sense of accountability and responsibility on the part of the decision makers.

In Section 4.1, we discuss the rather severe assumptions needed to derive the economic order quantity (EOQ). This is followed in Section 4.2 by the actual derivation. In Section 4.3, we show that the total relevant costs are not seriously affected by...