Part family and machine cell formation in multiperiod planning horizons of cellular manufacturing systems

Abstract

This paper presents a mixed integer mathematical model for simultaneous machine/part grouping and assignment when part-mix and demand changes across multiple periods according to forecasts and/or product life cycles. The model optimizes the trade-offs between the costs of cell configuration, machine procurement and salvage, part subcontracting, intercell movement, and capital investment. To improve the model's computational efficiency, a heuristic method of machine cell and part family formations is proposed and integrated with a mathematical program, which optimizes the various cost aspects. The heuristic is based on the criteria of maximum cell similarity and minimum number of machines. Experimentation with the proposed models shows that the heuristic procedure can be as powerful as the mathematical program, while being more numerically effective. Further experiments with the heuristic procedure shows its ability to arrive at improved solutions under varying system cost parameters.