Hierarchical production planning for multi-product lines in the beverage industry¶

¶Much of this work was done while at Vector Cons. Ltd.

Abstract

Multi-commodity production and distribution scheduling is one of the most complex and crucial problems facing many manufacturing companies. For a major European manufacturer specialising in bottling juices and drinks, we have designed and developed a hierarchical decomposition approach to the solution of the multi-commodity production planning problem. In this paper we focus our attention on the coarsest decomposition level, called multi-commodity aggregate production planning (MCAP). It concerns the choice of the best feasible production plan for a set of products (commodities) over an extended time horizon so as to meet forecast aggregate demands throughout the horizon. At this level, the problem constraints include hard constraints (such as production lines having a maximum capacity and products having short life-times), and soft constraints (budgetary concerns.) The objective is to determine the production plan that covers each period's demands as best as possible, while minimizing all relevant costs. Our method for solving MCAP produces optimal plans in negligible times in commodity PC workstations.

¶Much of this work was done while at Vector Cons. Ltd.

Keywords:

• Production planning
• Hierarchical decomposition
• Network flows
• Mathematical programming

Notes

¶Much of this work was done while at Vector Cons. Ltd.