On Designing Mixed Nonparametric Control Chart for Monitoring the Manufacturing Processes

TLDR: A modified nonparametric exponentially weighted moving average chart under progressive setup based on sign and arcsine test statistics is proposed and performs efficiently in detecting small and persistent shifts in the process location under each choice of the smoothing parameter.

Abstract: Efficient control charts are used to maintain and improve the manufacturing, industrial and service processes. Control chart is a basic tool of quality practitioners for producing products that meet market requirements in terms of quality, reliability and functionality. Lack of normality is a common occurrence in these processes. In such situations, distribution-free control charts are increasingly used for process monitoring. The nonparametric exponentially weighted moving average chart is a frequently used memory-type control chart in the process monitoring. The drawback of this chart is that it performs efficiently only on smaller values of the smoothing parameter. To compensate this drawback, a modified nonparametric exponentially weighted moving average chart under progressive setup based on sign and arcsine test statistics is proposed in this study. The prominent quality of the proposed scheme is that it performs efficiently in detecting small and persistent shifts in the process location under each choice of the smoothing parameter. The performance of the proposed chart has been investigated through simulations that use run length profiles (average run length, median run length and standard deviation of run length). When the performance of the proposed chart is compared with alternatives, its ability to detect small and persistent shifts is much better. Two real-life applications associated with hard-bake and piston rings manufacturing processes are included that show the demonstration of the proposed charts.