Control charts using minima instead of averages

Traditional control charts are commonly based on the averages of the inspected groups of observations. It turns out to be quite worthwhile to consider alternative approaches. In particular, a very good proposal is to use instead the group minimum for comparison to some suitable upper limit (and likewise the group maximum for comparison to a lower limit). The power of detection during Out-of-Control of the resulting chart is comparable to that of the standard Shewhart approach, while it offers much better protection to the effects of parameter estimation and/or nonnormality than the traditional methods.