7 Taguchi Methods (Robust Design) and Traditional Experimental Optimization Procedures Available to Purchase

Engineers in manufacturing industries face many challenges such as defects in production processes, warranty claims, troubleshooting issues, as well as constant changes in manufacturing conditions, which prevent productivity improvement. This chapter discusses the shortcomings of traditional product/process experimental optimization procedures compared to Taguchi Methods. Currently, the commonly used experimental optimization methods in manufacturing industries are one-factor-at-a-time or engineering judgment approaches. The one-factor-at-a-time method changes the setting of one factor to see its effect on the output while keeping the other factors constant. This approach is commonly used in academic and scientific research. Engineers use trends to identify optimal settings of experimental factors individually in order to optimize the output response. This one-factor-at-a-time approach does not consider interactions among experimental factors, although it is an easy method to follow. The observed individual optimal settings for experimental factors are seldom the true optimal combinations for the total system since interaction effects and energy transformations of the system are not evaluated. Another approach, the engineering judgment method, is used to resolve emergent manufacturing issues in current production processes. The engineering judgment approach usually takes two or three experimental runs to correct issues. Japanese and U.S. manufacturing industries use these approaches approximately 70% (one-factor-at-a-time) and 30% (engineering judgment) of the time.