Outlier‐Resistant Process Monitoring Using a RCV‐Based EWMA Control Chart

Traditional control charts usually fail to perform well in the case of outliers, or when the mean and variability of a process vary at the same time in quality control. In this paper, a very powerful exponentially weighted moving average (EWMA) control chart using the robust coefficient of variation (CV) has been proposed to overcome these hurdles. Unlike approaches that require data transformation or adaptive tuning, the scale‐invariance of the robust CV‐statistic can be used to stabilize the proposed RCV‐EWMA chart, keeping it sensitive in a polluted data setting. Through extensive Monte Carlo simulations, the capabilities of the chart are analyzed in different conditions of contamination levels and shift level. The total key performance indicators, such as Average Run Length (ARL), Standard Deviation of Run Length (SDRL), and Median Run Length (MRL) are given. The findings show that the suggested chart has good in‐control characteristics (ARL = 370) and is a worthwhile chart to detect small‐to‐still shifts despite up to 25% contamination. Its applicability in practice is also confirmed by a real data application that demonstrates how the chart can determine the abnormalities in the processes and be resistant to outliers. The results place the RCV‐EWMA control chart as an effective and dependable modern tool of monitoring processes especially those in very noisy or highly variable conditions.