Innovations in technology assist the manufacturing processes in producing high-quality products and, hence, become a greater challenge for quality engineers. Control charts are frequently used to examine production operations and maintain product quality. The traditional charting structures rely on a response variable and do not incorporate any auxiliary data. To resolve this issue, one popular approach is to design charts based on a linear regression model, usually when the response variable shows a symmetric pattern (i.e., normality). The present work intends to propose new generalized linear model (GLM)-based homogeneously weighted moving average (HWMA) and double homogeneously weighted moving average (DHWMA) charting schemes to monitor count processes employing the deviance residuals (DRs) and standardized residuals (SRs) of the Poisson regression model. The symmetric limits of HWMA and DHWMA structures are derived, as SR and DR statistics showed a symmetric pattern. The performance of proposed and established methods (i.e., EWMA charts) is assessed by using run-length characteristics. The results revealed that SR-based schemes have relatively better performance as compared to DR-based schemes. In particular, the proposed SR-DHWMA chart outperforms the other two, namely SR-EWMA and SR-HWMA charts, in detecting shifts. To illustrate the practical features of the study’s proposal, a real application connected to the additive manufacturing process is offered.

In this paper, single-phase mixture regression cum ratio estimators are presented by utilizing auxiliary variables and auxiliary attributes simultaneously under stratified random sampling. Special cases of these estimators are discussed and further mean square errors are extracted mathematically. Also, to observe the properties of proposed estimators, simulation technique is used which shows that the distribution of the proposed estimators is approximately normal. To differentiate the performance of the proposed estimators, an empirical study has been conducted by incorporating quantitative and qualitative characteristics in the form of auxiliary attributes and variables simultaneously. Comparisons are made with single-phase mixture regression cum ratio estimators under simple random sampling. It has been found that the mixture regression cum ratio estimators employing multiple auxiliary variables and attributes, simultaneously, under stratified random sampling are more efficient than mixture regression cum ratio estimator under simple random sampling.