Medium-Term Scheduling of an Industrial Multiproduct Batch

Plant

Esmael R. Seid, Jui-Yuan Lee, Thokozani Majozi

School of Chemical and Metallurgical Engineering, University of the Witwatersrand, Johannesburg,

1 Jan Smuts Avenue, Braamfontein 2000, Johannesburg, South Africa

Abstract

The pharmaceutical industry is one of the batch process industries where multiple high value products are produced with shared resources (e.g. equipment, raw material, manpower and utility) in the same facility. However, sharing of these resources makes planning and scheduling a challenging exercise. In the last two decades the research work focused on either long-term planning or short-term scheduling problems with little attention to medium-term scheduling.
This work presents a mathematical model for medium-term scheduling of an existing pharmaceutical industry that specializes in animal vaccines. The intractability of applying short-term scheduling models directly to medium-term scheduling problems has been overcome by adopting a decomposition strategy for the problem. The decomposition framework involves a two-level mathematical formulation. The first level determines which products and how much of each to produce in each scheduling subproblem to satisfy the market demand. The second level then determines the detailed sequencing of the tasks for the subproblems of tractable size. A recently developed robust short-term scheduling model based on a continuous-time representation was extended to solve the subproblems of the second level. The proposed model was successfully applied to a makespan minimization problem with a medium-term scheduling horizon of nearly 13 weeks.

Keywords: Medium-term scheduling; pharmaceutical industry; Multiproduct batch plant; Optimization.