Economic design of X control charts for monitoring batch manufacturing processes

The issue of economic design of an X control chart is addressed for a batch manufacturing process, subject to Markovian failure characteristics. The objective of this design is to minimise the expected total quality cost per batch. To find the optimal sample size and control limits, a Markov chain model is formulated. There is no analytical solution to the optimisation problem, so to avoid an exhaustive search, a heuristic approach, based on a regression model with the Cobb-Douglas function, is proposed. The impact of the model parameters and their interactions on the optimal solution is explored by conducting analysis of variance on 6561 numerical examples in a factorial design. Finally, an example of a real application demonstrates the effectiveness of the proposed model and heuristic.