Generic Models and Exact Algorithm for Process Capacity Maximization

ABSTRACT

In this paper, we study an optimization problem for maximizing a generic process's capacity in operations management. Process capacity is defined in this context as the maximum sustainable flow rate. The problem not only considers common features of processes such as collaboration, multitasking, batching, and setups, but also takes account of some features of the processes in which activities are sequentially or parallelly executed and also can be revisited. Unlike previous related analytical studies, we adopt mixed‐integer linear programming (MILP) to develop our decision models as well as an exact algorithm for capacity maximization of generic processes considering the aforementioned process features. The benefit of the MILP models lies in their easy extensibility by adding constraints, making them applicable to a broad range of real manufacturing or service operation processes with certain specific features. To shorten the decision time, the exact algorithm also contains some carefully designed cuts and an embedded dynamic programming module with dimensionality reduction and clustering. Managerial insights are obtained using the proposed methodology. For example, as the batch size of each activity increases, the process capacity grows, but as the number of units of each resource increases, the benefit of a greater batch size does not increase. Thus, operations managers should determine an appropriate batch size for each activity, as a larger batch size does not necessarily result in process capacity improvement. The results of this study also suggest that process capacity can be improved through making some, but not all, resources flexible.