A COMBINED ADAPTIVE SAMPLE-SIZE AND SAMPLING INTERVAL (X)OVER-BAR CONTROL SCHEME

A conventional XBAR control chart works in an open loop manner. That i s, the process control engineer waits until the control scheme signals an out-of-control condition. This paper proposes a combined adaptive XBAR chart that uses real-time, dynamic information available from the process to make the control scheme pro-active. The procedure consists of a simple method of varying the control chart design parameters (sa mple size and sampling interval) between minimum and maximum values de pending on the current state of the process. The proposed procedure is compared with the standard Shewhart XBAR chart, a variable sample siz e chart, and a variable sampling interval chart in terms of the averag e time to signal an off-target process. Substantial improvement in the performance over these schemes is demonstrated. Optimal values for th e maximum sample size are also given. The advantages of using the comb ined adaptive chart is illustrated by an example from a production pro cess.