Efficiency comparison of control charts for monitoring a positive mean shift in Poisson process

Abstract

Attribute control chart is a common statistical tools used to monitor number of defects or nonconformities in a manufacturing process, considered as count data and are Poisson distributed. The nonconformities are count data and are Poisson distributed. In this study we compare the performance of PCUSUM, PEWMA, ISRT c EWMA and PPM charts for detecting a positive mean shift in Poisson process. The plotting statistics are adjusted for monitoring a positive shift. The performances of control charts are evaluated by measuring the average run length when the process is out of control (ARL₁). The results reveal that the PCUSUM chart performs well to detect a small shift in Poisson process with low process average. PPM chart outperforms than the others for detecting small to moderate shifts, whereas PEWMA ( = 0.1) and ISRT c EWMA perform better than the others to detect a large shift.