AN OPTIMAL CONTROL CHART FOR NON-NORMAL PROCESSES

Abstract In this paper we present the application of a control chart for non-normal processes. This chart, which looks like an X/S control chart is built with a leastsquares L-estimator, which can replace the arithmetic mean and standard deviation usually calculated for Shewhart charts. This estimator has the property to provide a minimwn variance estimation of the process position and scattering. This, disregarding data distribution. We focused our attention on «multi-generators» processes, like screw-machines or multi-die holder for injection molding, these processes have the property to generate non-normally distributed pieces.