Contingency rules to control a production process subject to a single assignable cause

Abstract

We present two contingency rules to control a production process subject to a single assignable cause, i.e. the process shifts from in-control to out-of-control state when the single assignable cause occurs. In this paper, we assume a 100 per cent inspection policy as opposed to the sampling concept. However, the first rule we suggest is comparable with the traditional sampling model in that its criterion to intervene in the process is based on the number of defective products. The second rule uses more information than the first does: it triggers intervening in the process when the inter-arrival time between two consecutive defective products is smaller than an optimally derived cut-off rate. We first show how the two rules can be derived, and, as a preliminary analysis, propose conditions under which one contingency rule might be more effective than the other.