Improved adaptive CUSUM control chart for industrial process monitoring under measurement error

Abstract

Measurement error (ME) is a critical factor that affects the accuracy and reliability of statistical process control (SPC) methods, often leading to delayed fault detection and compromised process monitoring. This study proposes an improved adaptive cumulative sum (IACUSUM) control chart that effectively mitigates the adverse effects of ME by integrating a linear covariate model and a multiple measurement procedure. The performance of the proposed chart is evaluated using average run length (ARL) and standard deviation of run length (SDRL) through rigorous Monte Carlo simulations and real-data applications. The findings demonstrate that ME significantly impacts the detection capability of control charts, underscoring the need for effective error management strategies. The IACUSUM control chart, when implemented with a multiple measurement approach, exhibits superior sensitivity, enhanced shift detection, and greater robustness compared to conventional methods. The results confirm that the proposed methodology significantly improves process monitoring efficiency, making it a highly reliable tool for industrial applications where measurement variability is prevalent. This study provides a practical and scalable solution for enhancing SPC performance and sets the foundation for further advancements in adaptive control charts for real-world quality assurance systems.