Track: Quality Engineering, Control and Management

Abstract

Exponentially weighted moving average (EWMA) chart is one of the most powerful techniques for detecting small and moderate shifts. The curtailed inspection has been broadly used in acceptance sampling plans to reduce the average sample number substantially. This research presents an EWMA chart employing curtailed inspection (curtailed EWMA chart) to monitor the fraction nonconforming p of a process with attribute characteristics. A design algorithm is developed to optimize the charting parameters of the curtailed EWMA chart. The proposed curtailed EWMA chart is compared with the traditional EWMA without curtailed inspection (abbreviated as EWMA chart in this research) using the same false alarm rate to ensure a fair comparison. The overall detection ability of the charts is measured in terms of the expected average run length over a wide range of shifts, and the p shift is assumed to follow a uniform distribution. The findings of this research reveal that the curtailed EWMA chart has a superior overall performance than the EWMA chart without curtailed inspection. On average, the former is more effective than the latter by 35%, considering different circumstances. The high overall effectiveness of the curtailed EWMA chart is mainly attributable to curtailed inspection.