Impact of scheduling frequency and shared capacity on production and inventory costs

Abstract

A multi-product production planning problem is considered. Products can either share the same production line, or be produced on different lines. We study the influence of planning cycle length on inventory and production performance, first in terms of how these affect the variability of the states in the production system, and then how this translates into cost. Production costs are assumed to be dominated by labour costs, where we use a model with guaranteed hours and overtime. Inventory performance is measured using holding and backlog costs. The replenishment policy used is the Proportional Order-Up-To (POUT) policy, containing as a special case the classical OUT policy. It does this by having a controllable feedback parameter, which alters the relative dynamics of inventory and capacity. Demand is assumed to be normally, independently and identically distributed. Our analytical results show that there exists an optimal rescheduling frequency that minimises total cost. This frequency is determined by the relationship between the costs and lead time. The economic benefit from planning in larger buckets follows from demand pooling, which allows more effective use of guaranteed hours and reduces over-time. A balance must be struck between reducing inventory costs, which generally get smaller as we plan more frequently, and labor costs, which increase as we plan more frequently. We also demonstrate that the pursuit of ever-faster rescheduling leads to severe cost penalties. A comparison between the benefits of optimizing the rescheduling frequency, production line consolidation, and POUT parameter tuning highlights the economic potential of optimizing rescheduling frequency.

Publication
Pre-prints of the 17th International Working Seminar of Production Economics, Innsbruck, Austria, 20th-24th Feburary, Vol. 2, 277-288