Process dispersion monitoring: Innovative AEWMA control chart in semiconductor manufacturing

Abstract

This study introduces a novel adaptive exponentially weighted moving average (AEWMA) control chart for monitoring process dispersion, employing adaptation to determine the smoothing constant. It is designed to effectively track shifts within expected ranges in process dispersion by computing the smoothing constant through a suggested adaptive approach. To determine its efficacy, the chart’s performance is evaluated by using smaller run-length profiles derived from Monte Carlo simulations. A key feature is the utilization of an unbiased estimator to calculate the smoothing constant via the proposed function, enhancing the chart’s ability to detect various magnitudes of decreasing and increasing process dispersion shifts. A comparison with an existing adaptive EWMA dispersion chart underscores the significant efficiency of the proposed chart across various types of process dispersion shift magnitudes. Moreover, the study encompasses a real-life dataset application, demonstrating the practical implementation and ease of use in real-world scenarios.