A single-period inventory placement problem for a supply system with the satisficing objective

• Published in: European journal of operational research Vol. 224; no. 3; pp. 520 - 529
• Main Authors: Chung, Chia-Shin, Flynn, James, Kuik, Roelof, Staliński, Piotr
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 01.02.2013; Elsevier; Elsevier Sequoia S.A
• Subjects: Decision analysis; Demand; Inventory; Inventory management; Newsvendor; Optimization algorithms; Product returns; Salvage value; Satisficing; Single period; Stochastic; Studies; Supply chain; Supply chains; Single period; Supply chain; Satisficing; Stochastic; Newsvendor; Inventory
• ISSN: 0377-2217; 1872-6860
• DOI: 10.1016/j.ejor.2012.09.009

► N stocks meeting random demand in one period are considered. Unmet demand is lost. ► Economics, such as set-up, holding & purchase costs, can vary across stocking points. ► Stock is allocated maximizing the probability of attaining a fixed target profit. ► We find that one only needs to utilize three stocks in an optimal allocation. ► We present an efficient algorithm finding an optimal solution with three stocks. Consider the inventory placement problem in an N-stage supply system facing a stochastic demand for a single planning period. Each stage is a stocking point holding some form of inventory (e.g., raw materials, subassemblies, product returns or finished products) that after a suitable transformation can satisfy demand. Stocking decisions are made before demand occurs. Unsatisfied demands are lost. The revenue, salvage value, ordering, transformation, and lost sales costs are proportional. There are fixed costs for utilizing stages for stock storage. The objective is to maximize the probability of achieving a given target profit level. We prove the existence of optimal stocking decisions where at most three stages receive nonzero stocks. We also characterize properties of the optimal stocking decisions and provide an O(N3) algorithm for their computation. For the special case where all fixed costs are zero, the stages utilized do not depend on the demand distribution or the target level, and one can find optimal stocking decisions by performing a simple O(N2) search and solving a single-variable optimization problem.