A Group of Weighted Moving Averages: An efficient plan for monitoring a range of process changes

Abstract

It is common to monitor quality characteristics of process outputs or inputs in production processes. Early detection of deviations from target is an important part of the intervention strategy for keeping the process on target. This paper proposes the use of moving averages with smaller sample sizes to provide an early signal of large location changes from target, while moving averages with larger sample sizes are used to give an early signal of smaller location changes from target. If the averaging process gives more recent observations greater weight then it responds to changes quicker than the simple average. Therefore, this paper illustrates that the joint application of a group of weighted moving averages of variable sample sizes is an efficient monitoring plan for early detection of both large and small deviations from target. These plans are compared to the conventional cumulative sum (CUSUM) and exponentially weighted moving average (EWMA) plans using simulation studies and two applied examples.