This research develops improved statistical methods for both planning and analyzing robust design experiments. A new experiment format, the combine array format, will be developed to reduce the experiment size and allow greater flexibility for estimating effects which may be more important for physical reasons. Design strategies, alternative graphical tools and tables, and computer algorithms will also be developed to help engineers plan more efficient experiments themselves. For analyzing experiments, the response model approach will be developed to yield additional information about how control parameter settings dampen the effects of individual noise parameters. Alternative variability measures for Taguchi's signal-to-noise ratios and methods for empirically determining the appropriate measure to use will also be developed. The impact of this research is expected to be significant in terms of expanding the science base of techniques for improving product quality, manufacturability and reliability at low cost.
