A manager confidence oriented design scheme of control charts

Manufacturing is a process of realization of products. Product reliability, durability and other characteristics are determined in manufacturing and will degenerate because of process variation. Statistical process control charts have been an effective tool to help alarm process shifts. A proper design scheme is important because poor parameters will lead to poor power of control charts which will greatly affect manager confidence on SPC monitoring. In most economic design schemes, the control chart parameters are chosen in such a fashion that the expected cost of monitoring is minimized. Though there have been plenty literatures, few applications of economic designed control charts could be found in practice. One barrier is that managers have low confidence on the designed parameter. Usually only minimizing the expectation of cost does not reflect the manager preference sufficiently. Risk aversion exists whilst exposed to uncertainty of cost. Also false alarms make control charts mistrustful. In this paper, oriented to building up manager confidence, the design scheme is reconsidered. A multi-criteria decision making model with objectives of low cost, low risk and low frequency of false alarms is proposed to conduct the new design scheme. Pareto optimality and data envelopment analysis technique are applied to aid making decision. Cases are established to illustrate the validity of this new scheme.