A new fuzzy-c-means and assignment technique-based cell formation algorithm to perform part-type clusters and machine-type clusters separately

The incorporation of fuzziness in the cell formation problem by Chu and Hayya is a notable contribution, in which nonbinary classification logic is used. However, despite this development numerical illustrations performed in this research demonstrate that the Chu and Hayya approach can result in solutions with empty part-type-cluster(s) and/or empty machine-type-cluster(s). Further, it is noted that solutions based on the Chu and Hayya approach can contain non-empty part-type-cluster(s) being assigned to empty machine-type-cluster(s) and vice versa. Three strategies are offered in this research to overcome these inadequacies: the separate formation of part-type- and machine-type-clusters, modification of the stopping criterion, and the adoption of the assignment technique in the formation of the final manufacturing-cell solutions. A new algorithm has resulted from these modifications and have been rigorously compared to the performance of the Chu and Hayya approach. In general, the new algorithm demonstrates superiorities in global efficiency and in generating feasible part-type-clusters, machine-type-clusters and manufacturing-cells.