On the enhanced surveillance methods for high-quality processes

Nowadays, processes are equipped with advanced tools; therefore, they produce near zero-defect items and are termed as high-quality processes. The high-quality data often follows the Zero-Inflated Poisson or Negative Binomial (ZIP or ZINB) distributions. In literature, most surveillance methods are designed to monitor ZIP and ZINB distributed quality characteristics. However, some covariates are also available along with the count-based quality characteristics of a process. Therefore, this study is intended to propose moving average (MA) and double MA (DMA) based surveillance methods designed on the ZIP and ZINB residuals (i.e., Pearson). A simulation-based study is carried out to evaluate the performance of proposed methods and their comparative results with an existing method using run-length properties. The findings reveal that the proposed MA and DMA methods outperformed the existing Shewhart chart. Moreover, a real-life example is presented, which supports the simulated results.