MEWMA Chart

The Multivariate EWMA (MEWMA) control chart is a statistical process monitoring method designed to detect small and sustained shifts in the mean vector of a multivariate process. Introduced by Lowry, Woodall, Champ, and Rigdon in 1992, it extends the univariate EWMA chart to p-dimensional observation vectors by computing an exponentially weighted moving average of successive measurement vectors and charting a Hotelling-type quadratic form against a control limit determined by a target average run length.