Robustness Indices in Multistage Problems.

The study of robustness in single-stage manufacturing has been surveyed by many researchers, while modeling of robustness in multistage manufacturing using multivariate data is seldom explored. The aim of this paper is to develop a methodology to model process robustness in a multistage manufacturing system. This paper proposes a methodology to study the robustness of a multistage process that creates a component and process capability that acts as a surrogate measure of robustness. A powerful technique to facilitate the analysis of the mentioned problem, response surfaces express the behavior of responses considering the above controllable factors (quality characteristics). A common approach to estimating response surfaces using experimental data is the ordinary least squares method. Because this method is very sensitive to outliers and correlated responses, in this paper, a robust coefficient estimation method is proposed for multiresponse surfaces in multistage processes based on M-estimators. The results show that considering the proposed robustness criteria based on a multivariate process capability index, the robust multistage process would perform better than the individual stage process. This paper adds to the body of knowledge on the robustness of multistage manufacturing processes, and also the relationship between robustness and capability is to be explored as surrogate robustness indices in the multistage process. The efficiency of the proposed approach is assessed by a small simulation study besides a real manufacturing case study. Copyright © 2017 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]