Improving profitability of optimal mean setting with multiple feature means for dual quality characteristics

Abstract

The setting of a process mean for a manufacturing process which frequently produces scrap and rework can significantly affect profitability. Optimal mean setting is a methodology by which the process mean is adjusted to maximize profit. This paper studies the dynamics of the problem and investigates the possibility of applying different process means to each rework iteration, to further maximize profit. A proof is given confirming there is only one optimal mean that applies over all rework iterations in the single feature case. However, applying similar reasoning to a dual feature case led to the development of a new optimal mean setting methodology which outperformed the existing approach in terms of the maximum expected profit.