Improved EWMA and CUSUM Charts Under Modified Successive Sampling for Monitoring Process Dispersion

Abstract The Statistical Process Control (SPC) toolkit is extensively utilized to identify variations in processes, with control charts serving as the most efficient and commonly employed instrument for real-time process monitoring. Control charts evaluate whether a process is stable or unstable, detecting special cause fluctuations. Monitoring process variability is generally prioritized over location characteristics. Although quality evaluation samples are typically obtained via simple random sampling (SRS), the modified successive sampling (MSS) method is favored to reduce sampling duration and expenses. This research formulates CUSUM and EWMA control charts employing the MSS methodology to assess process variability. Performance criteria, such as run length measurements, are employed to evaluate the efficacy of CUSUM and EWMA charts in comparison to Shewhart charts. The results demonstrate that the EWMA chart surpasses both the Shewhart and CUSUM charts. A practical illustration from fertilizer production is provided to exemplify the proposed methodology.