Abstract (Translated):
Multivariable capacity index have been used frequently in the manufacturing industry and similar environments, since they provide quantitative measurements of the potential and performance of a process described by multiple quality characteristics susceptible to evaluation and correlated simultaneously. In this paper, we present an empirical evaluation of several approaches to analyze a multivariate industrial process. In the firs approach we study relationship of the volume of the tolerance region and the volume of the process region. In the second approach, we propose the application of principal components technique and, finally, we propose the application of other indicators that analyze the proportion of observations located outside the specification margins and their variability in the long and short term. The results provided indicate that analyzed production process is not marginally capable for satisfying predefined technical specifications and that there is a high margin of improvement.