Browse Journals

Recent applications of control charts show that they are capable of monitoring not only manufacturing processes but also service processes. Data sets generated by several service processes often come from a process with non-normal distribution or unknown distribution. As most of the existing charts are not suitable to monitor such non-normal processes, in this paper, a new type of ‘Cumulative Sum (CUSUM)’ chart based on arcsine transformation over a simple statistic is proposed to monitor the shifts of the process mean of a few service processes. Using the sampling properties of the new monitoring statistic, the average run lengths of the proposed chart have been calculated to check if the chart (i) is capable of detecting both small and large shifts in the process mean and (ii) performs as good as the transformed EWMA (exponentially weighted moving average) chart due to Su-Fen et al. ([9]) to monitor processes that exhibit a drifting mean over time. A numerical example of service times with skewed distribution from a service system of a bank in Taiwan is used to illustrate the performance of this chart.

in-control, arcsine, cusum, statistical process control, nonconforming.