Monitoring process variability using decile mean standard deviation

Abstract

AbstractThe control charts are essential tools in order to monitor the process quality as well as are used in numerous industries. The classical Shewhart S‐control chart is the most prevalent control chart in Statistical Process Control (SPC) for assessing changes in process over time. This paper modifies the S‐control chart by proposing SDDM‐control chart based on the SDDM (decile mean standard deviation), which presume that the quality function is monitored normally. The traditional and proposed control charts’ performances under study are evaluated via a Monte‐Carlo simulation study by estimating expected widths (EW) as well as expected out‐of‐control points (EPO). In the simulation, the average out of control run length (ARL1) is considered the best method for analyzing sensitivity of the control charts. The outcomes of the simulation analysis revealed that the proposed SDDM‐control chart is well performed and very close contender of the classical S‐control chart insofar that the improvement in process variability is very similar to that of the S‐control chart. In addition, the architecture layout of the proposed SDDM‐control chart is more powerful than the conventional S‐control chart for the smaller sample size value (n) that is the most realistic situation in SPC applications. Hence the proposed chart is equally compatible for shift detection under normal process. In case of abnormal processes, the proposed SDDM‐control limits are less affected by the presence of outliers. A real‐life dataset is evaluated to help our observations from the simulation analysis for illustrative purposes. Consequently, the SDDM version of the S‐control chart is recommended to be used by the practitioners in various fields of industry, engineering, medical and physical sciences.