Modeling multiple plant sourcing decisions

Abstract

With the recent trend of company acquisitions and mergers, it is likely that the company that results from a merger will own multiple facilities with overlapping manufacturing capabilities. By shifting the production of parts to different plants, capacity can be better utilized, and the overall cost to the company can be reduced. This paper addresses the sourcing decisions of allocating parts to several internal facilities or to an external supplier. This multiple-plant sourcing problem (MPSP) is modeled as a multiple-choice multidimensional knapsack problem that considers the capacity of each facility and the costs associated with supplying a part at each facility and externally. We present results from an analysis of two-, three-, and four-facility MPSP models and find that the two-facility model is the most difficult to solve. Additionally, we model extensions that incorporate allocating a part to more than one facility with various inventory strategies, and consider how these extensions affect the sourcing decisions.

Keywords:

• Sourcing decisions
• Multidimensional knapsack problem
• Multiple-choice multidimensional knapsack problem
• Multiple plants

Acknowledgements

The authors are appreciative of the constructive comments of two referees, especially for directing us to the multiple-choice multidimensional knapsack problem literature, considering this literature provided insights into the results presented in section 4.