Designing robust products with multiple quality characteristics

This study presents an approach for designing robust products with multiple quality characteristics. The development of the proposed approach is based on Taguchi's concept. However, the robustness is improved in terms of the designers' degree of satisfaction. The approach first determines the transformation function of satisfaction based on the degree of importance for each quality characteristic. Once the S/N ratios are obtained for each quality characteristic, they are transformed into the degrees of satisfaction using a transformation function. Using the degrees of satisfaction as the response, the control factors for each quality characteristic are found, and a corresponding multiple regression model is fitted. Then a mathematical programming problem, consisting of the associated models, is formulated. A set of optimum settings that maximizes the designers' overall satisfaction is then determined. Alternatively, the approach allows the designers to achieve the aspiration levels of satisfaction for some quality characteristics by simply adding the associated constraints to the problem. An example is used to demonstrate the applicability of the proposed approach.