could be jointly monitored using a multivariate control chart. Points that lie outside the control limits on these charts would identify periods during which the control errors are large or
during which large changes to the manipulated variable are being made. These periods
would likely be good opportunities to search for assignable causes. Montgomery ct al.
(1994) have demonstrated the effectiveness of such a system. Other very useful references
and discussion are in Del Castillo (2002). Figure 12.13 illustrates how such a combination
of EPC and statistical process monitoring might be employed.