Abstract
<title>Abstract</title> <p>Statistical Process Control (SPC) often struggles to efficiently monitor both minor and major process shifts within a single framework. To address this challenge, this study introduces a Dynamic Huber-Weighted Adaptive Exponentially Weighted Moving Average Max-Multivariate (AEWMA Max-M) statistic designed for simultaneous multivariate process monitoring. By integrating a Huber-based scoring function, the proposed scheme dynamically adjusts its smoothing parameters based on the magnitude of observed deviations. For subtle changes, the framework heavily weights historical data to enhance sensitivity, whereas for significant errors, it reduces the influence of past data to rapidly detect major structural shifts. The diagnostic performance of this chart was rigorously evaluated using Average Run Length (ARL) simulations. The findings demonstrate that the proposed dynamic framework systematically outperforms the traditional EWMA Max-M and basic Max-M models in efficiently pinpointing isolated and concurrent shifts in the process mean and variance. Furthermore, the practical efficacy of the AEWMA Max-M scheme was validated using multivariate industrial quality data from cement clinker production across five distinct parameters. The empirical application confirmed its superior responsiveness, successfully flagging a maximum of ninety-eight distinct out-of-control instances under high-sensitivity configurations. Ultimately, this adaptive statistic provides a robust and versatile tool for complex quality surveillance.</p>