No abstract

A centralized inventory system serves a number of stores with common ownership, and thus reliable and timely information sharing. Each of them pays a share of the inventory cost, and the reward structure leaves the owners of individual stores rewarded for their individual performance. Appropriate selection of a cost allocation method is important if such a centralized system is to last. In this work we propose three necessary criteria—stability (core of a related cooperative game), justifiability (consistency of benefits with costs), and polynomial computability. For a concrete example we demonstrate that common allocation procedures may not meet all three tests, and we present a method that that meets all three criteria. This kind of cost allocation analysis helps the common management to evaluate the trade‐offs in choosing an allocation scheme for the cost of inventory centralization. © 1996 John Wiley & Sons, Inc.

No abstract

This paper examines the subject of cost allocation in a multiple product inventory system, allowing for consolidation of shipments. If we order multiple items using an economic order quantity (EOQ) policy, and consolidate shipments, part of the ordering cost is shared, and part is specific to each item; we want to find the consolidation choice with optimal total cost and divide the cost fairly among the individual items. Such a fair division is central to a costing system in which no group of items subsidizes the others; there are no free riders! We use a cooperative inventory game to determine when this can be done. This game is usually not concave, so we want to know what consolidation combinations determine when this cost can be fairly divided, using the core of the game. We prove that consolidation of all the items is cheaper exactly if there are fair cost allocations (core of the game is not empty), which happens when the portion of the ordering cost common to all items is not too small. We further show how sensitive the nonempty core result is to adjustments in the cost parameters and show how to determine a threshold value for the shared ordering cost, which assures the existence of a fair cost allocation.inventory games, cost allocation, shipment consolidation

No abstract

Agent technology offers a new way of thinking about many of the classic problems in operations research. Among these are problems such as project scheduling subject to resource constraints. In this paper, we develop and experimentally evaluate eight agent-based algorithms for solving the multimode, resource-constrained project scheduling problem. Our algorithms differ in the priority rules used to control agent access to resources. We apply our approach to a 51-activity project originally published by Maroto and Tormos [I]. We solve the problem using two types of agent-based systems: (i) a system of simple, reactive agents that we call basic agents; and (ii) a system of more complex, deliberative agents that we call enhanced agents. Of the eight priority rules tested, we find that priority based on shortest processing time performs best in terms of schedule quality when applied by basic agents while the priority based on earliest due date performs best when applied by enhanced agents. In comparing agents across priority rules, we find that enhanced agents generate much better schedules (with makespans up to 66% shorter in some cases) and require only slightly more computation time.

We examine cooperative games in supply-chain management termed Inventory Games. Supply-chain management has non-cooperative and cooperative interactions between the participating players. We provide a concise survey of cooperative inventory games in the form of extensions on two basic problems. For deterministic games, Economic Order Quantity-like policies with joint replenishment are of primary interest. For stochastic games we examine Newsvendor-like centralization games and their extensions. We conclude with a short summary of a dynamic Newsvendor realization game and directions for further research.

For single period inventory models with normally distributed, correlated individual demands we examine the problem of minimizing the cost of inventory centralization as a function of the covariance matrix. In a stable centralized setting there are no incentives for any party to break-away -referred to as nonempty core conditions. For the allocated benefits in inventory centralization, nonempty core conditions are always satisfied. In this paper we discuss a step by step greedy optimization procedure which computes an optimal centralization solution. The procedure manipulates the correlations without changing the mean or the variance at each store. We do not just accept that in the centralized setting the parties are better-off, but for the first time provide the analysis of how to maximize their collective benefits.