Optimal process variance under Taguchi loss

Most research on the process control charts concentrates on the economic design of its parameters. Reduction of variance as a control decision has not been researched. Presents a model for optimal decision on variance reduction and includes the rejection losses of the non‐conforming units increasing due to increased variance, Taguchi loss of the conforming units, and the cost of reducing variance. Optimal policies are derived analytically for uniform distribution and numerically for normal distribution. Applications of the model to the area of machine replacement and global manufacturing are suggested.