Application of Principal Component Analysis in Machine-Part Cell Formation

TL;DR

This paper demonstrates how Principal Component Analysis (PCA) can effectively group machines and parts for cell formation, improving efficiency in manufacturing layouts.

Contribution

It introduces a PCA-based methodology for machine-part cell formation and validates its effectiveness across multiple datasets compared to existing methods.

Findings

70% of cases show increased grouping efficacy

30% of cases match the best literature results

Method is validated on multiple machine-part matrices

Abstract

The present paper applied Principal Component Analysis (PCA) for grouping of machines and parts so that the part families can be processed in the cells formed by those associated machines. An incidence matrix with binary entries has been chosen to apply this methodology. After performing the eigenanalysis of the principal component and observing the component loading plot of the principal components, the machine groups and part families have been identified and arranged to form machine-part cells. Later the same methodology has been extended and applied to nine other machine-part matrices collected from literature for the validation of the proposed methodology. The goodness of cell formation was compared using the grouping efficacy and the potential of eigenanalysis in cell formation has been established over the best available results using the various established methodologies. The result shows that in 70% of the problem there is increase in grouping efficacy and in 30% problem the performance measure of cell formation is as good as the best result from literature.