A mathematical programming approach to evaluating alternative machine clusters in cellular manufacturing

Cellular manufacturing systems achieve the economies of scope and scale approaching that of flexible and high-volume production when the machine/part clusters are totally independent of each other. However, most real systems contain bottleneck machines and exceptional parts (exceptional elements) that reduce these economies. Many grouping methods have been proposed for creating the initial machine/part cells where the presence of exceptional elements may greatly affect their performance. Furthermore, multiple alternative solutions are often possible for a given grouping algorithm. In this paper, the previous work dealing with exceptional elements is reviewed. A mathematical programming model used for comprehensively dealing with exceptional elements is investigated. The effect of alternative initial machine/part clusters on the total cost is evaluated. It is demonstrated that the mathematical programming model can provide useful information in making trade-off decisions when exceptional elements are present.