Quantitative evaluation of Employee Substitutability in manufacturing using Cosine Similarity and Principal Component Analysis

Abstract

Employee substitutability in manufacturing plays a key role in ensuring production efficiency and seamless workflow transitions. This study explores algorithms and computational methods to quantify substitutability using operational performance data. In work are used employ Cosine Similarity and Principal Component Analysis (PCA) to evaluate how closely employees' work patterns align. Cosine Similarity measures the resemblance between employee task performance, while PCA reduces dimensionality to highlight key differences in skill sets. A dataset containing employee operation counts and execution times was analyzed using these methods. The results indicate that employees with high Cosine Similarity scores exhibit comparable performance levels, while PCA effectively identifies clusters of employees with similar efficiency patterns. Combining both techniques provides a comprehensive methodology for employee substitutability, workforce optimization and facilitating task reallocation.