Performance Analysis and Evaluation of Assemble-to-Order Systems With Non-Stationary Demands

Abstract

Assemble-to-order (ATO) systems, which keep inventory only for components and assemble the end products after demand is realized, have become prevalent in many industries. We study multi-products and multi-components ATO systems with stochastic lead times. Independent Base Stock (IBS) policies and First-Come-First-Serve (FCFS) allocation rules are used to control component inventories. To find the optimal inventory control parameters for such systems, it is necessary to first characterize the relationship between the system performance measures such as order-based backorders and order fulfillment rate and the inventory control parameters such as reorder points. Most previous studies have focused on ATO systems where the demands are homogeneous Poisson processes. In this paper, closed-form expressions are derived for the relationship between the above performance measures and inventory control parameters where the demands are non-homogeneous. We also examine the impact of common component scarcity and demand variability. Finally, the above conclusions are illustrated by numerical examples.

Keywords

• Assemble-to-order
• non-homogeneous Poisson process
• generating functions