Combination of two Shewhart-type control charts to detect small as well as large shifts in multivariate process variability

This paper proposes a procedure based on two Shewhart-type control charts for monitoring the stability of multivariate process variability. That procedure will be able to detect small shift as well as large shift in covariance structure. For that purpose, generalized variance and vector variance will be used as multivariate variability measures. If generalized variance chart is sensitive to the large shift, we show that vector variance chart is sensitive to the small shift. A real industrial example will be presented to illustrate the advantage of the proposed procedure.