A quality monitoring and decision-making scheme for automated production processes

Abstract

Traditional attribute control charts are based on the monitoring of the number of nonconforming items in a sample of fixed size. In modern manufacturing processes, since items can be checked automatically, use of samples of a fixed subjective size with traditional charts is not suitable for on‐line continuous inspection. Furthermore, the sample size usually has to be large when the process fraction nonconforming is not reasonably high. If a decision is to be made only when a sufficient number of items are manufactured and inspected, many nonconforming items might have been produced. In this paper, a control scheme is presented based on the monitoring of cumulative counts of items inspected. This procedure will limit the number of consecutive nonconforming items to a small value when the process has suddenly deteriorated, can detect a sudden process shift quickly and is suitable for continuous inspection processes. The exact probability control limits are derived and the implementation procedure is described.