A mean-shift pattern study on integration of SPC and APC for process monitoring

To detect a long-term mean shift of an autocorrelated process, traditional Statistical Process Control (SPC) techniques can be applied to monitor a process with Automatic Process Control (APC) or Engineering Process Control (EPC). In this paper, we investigate the relationships between the run-length performance, the mean-shift pattern, and the autocorrelation structure of the original process. For both monitoring the output and monitoring the control action of the APC-controlled process, we study how the mean-shift pattern affects the run-length distribution of the monitoring process. We compare the performance of the two monitoring approaches and make recommendations for various autocorrelated processes. We find that one can indicate the average run-length performance of an automatic-controlled process by examining the mean-shift pattern of the monitoring process.