An unbiased function-based Poisson adaptive EWMA control chart for monitoring range of shifts

For monitoring the number of nonconformities per unit in industrial processes during the inspection, Poisson control charts are most widely deployed. These charting structures are referred to as attribute control charts as the quality characteristic under study is based on a nominal scale other than a quantitative or measured scale. In this study, Poisson adaptive exponentially weighted moving average (PAEWMA) has been designed and the performance of the said chart under a steady-state situation along with the zero-state condition has been studied for detecting changes over an unknown range of shifts. The enactment of the proposed PAEWMA design with existing schemes has been evaluated and compared using run-length (RL) profiles such as; average RL (ARL), the standard deviation of RL (SDRL), and some percentile points of the RL distribution. The extra quadratic loss and relative mean index measures have been used for comparing the range of shifts. The comparison reveals that the proposed PAEWMA chart under the steady-state situations performs far better than the competitors. The effect of estimation of the process parameter is also a part of the study. The three applications of the proposal have been included; which are taken from the artificial dataset, from past study monitoring dataset, and the other from the aircraft accident monitoring data for the implementation of the proposal.