Abstract: | There has been considerable debate over the contributions made by Genichi Taguchi to robust process and product design. As a result of the numerous debates, there have been many alternative approaches presented that are better suited to the robust design problem. In this paper a combined array design is presented as an alternative to a standard Taguchi design. The mixed resolution design is illustrated in an example involving control and noise variables. Two new variance properties of experimental designs are also presented. Copyright © 2000 John Wiley & Sons, Ltd. |