Control Charts monitoring product’s loss to society

Taguchi introduced a new philosophy in quality control which accounts for the economic loss associated to process variation while considering deviations from the target value of a product quality characteristic. Taguchi loss function has been considered in the design of control charts only for the computation of costs associated to nonconformities. This paper considers sample statistics based on Taguchi loss function to implement Shewhart control charts monitoring both the deviation from the target and dispersion of normally distributed quality characteristics. The aim of the proposed control chart is performing on-line quality control of a process by monitoring its quality loss cost performance vs. time. To compute the quality loss performance we consider a nominal-the-best quality characteristic. The statistical performance of the proposed control charts has been evaluated and compared to that of widely used control charts. Implementing target costing philosophy by means of one of the proposed charts is also discussed. An illustrative example shows the Taguchi control chart practical implementation.