Fuzzy logic concepts applied to machine–component matrix formation in cellular manufacturing

Conventional cell formation procedures assume that each part feature is developed on a single machine. In many realistic situations, however, alternate machines are suitable for developing the same part feature. This is dependent on the uncertainties in the dimensional machine tolerance and processing times. Also, each machine can develop more than one part feature during a single operation. In addition, the development of a certain part feature is dependent on the preceding part features. In this paper, fuzzy logic methodology is employed to deal with these uncertainties and handle the dependency or interaction between part features during the determination of the overall suitability of a machine to process a part. The concept of linguistic hedges or modifiers is used to indicate the relative importance of a feature during this exercise. The resulting machine suitabilities are depicted in the form of a non-binary machine–component matrix. An example is also solved to illustrate our approach.