Probabilistic robust design of multiple response systems for functional design for Six Sigma

A relatively new concept related to Six Sigma, called Design for Six Sigma (DFSS), appears to have strong resemblance to robust design methods. DFSS is a means of developing, or improving, products that enable Six Sigma levels of performance in production while focusing on customer satisfaction and robustness. In this paper the means and tolerances of the design variables are determined (i.e. functional deign) such that the defect rate of multiple response systems satisfies, for example, a Six Sigma level goal. A constrained optimization approach is applied using production costs as the objective to be minimized and probability of non-conformance as the constraint. The proposed method is used to design both a voltage divider and a high pass filter. Monte-Carlo simulation is used to verify the results. Design using probability constrained optimization can be a promising method for DFSS.