Abstract
In this paper, we demonstrate some numerical results for the optimal economic and economic statistical design of exponentially weighted moving average (EWMA) control chart. The system is designed to operate with preventive maintenance policy because relatively few systems are designed to operate without maintenance of any kind. Hence we assume the process-failure mechanism obeys a Weibull-PM distribution instead of the traditional exponentially or Weibull distribution. We choose the constant length of sampling intervals by using the method that Banerjee and Rahim (1993) proposed. The computational results indicate that the Weibull-PM model provides a lower cost than Weibull model. They also show that the economic statistical design has higher cost than the pure economic design, but provides better in-control ARL. The sensitivity analysis of ARL bounds shows that small size of the shift has the sensitivity of cost for larger ARLu.