A Practical View of the Complexity in Developing Master Production Schedules: Fundamentals, Examples, and Implementation

Abstract

Although Master Production Scheduling (MPS) has been studied and used by both academia and industries for quite a long time, the real complexity involved in making a master plan when capacity is limited, when products have the flexibility of being made at different production lines, and when performance goals are tight and conflicting, has not yet been presented in the literature in a simple and practical way. In this context, one should consider how to attain a given performance by balancing different objectives, such as maximizing service level, and minimizing inventory levels, risk of stockouts, overtime, and setup time. Many decisions need to be made during the development of an MPS, such as: Which product should be scheduled, in what quantity, and to which resource? Is overtime needed? Should inventory be built for future periods? Should backlogging be considered? Clearly, an MPS process depends on the combination of many different parameters. For this type of problem, it is extremely difficult to find a solution that satisfies all objectives involved simultaneously, mainly because of the great number of variables involved. It is known that finding an optimal MPS solution for industrial scheduling scenarios is time consuming - despite nowadays computers being extremely fast. It is common, therefore, to use heuristics (or meta-heuristics) to find good plans in reasonable computer time. Using a plain language, this chapter describes some of the complexity involved in the MPS creation without, however, paying too much attention to mathematical formalisms and definitions, using mostly the author’s industry experience and practical examples faced during research in the production scheduling area.