A Comparison of Taguchi and Traditional DOE Approaches Using Response Surface Methodology: An Automotive Case Study

Walden, C., & Miller, J. (2005). A Comparison of Taguchi and Traditional DOE Approaches Using Response Surface Methodology: An Automotive Case Study. CAVS Extension Technical Report. Canton, MS.

Abstract

A manufacturing operation is a dynamic, integrated, and complex system; numerous potential factors drive performance. Experimentation with an actual production system is frequently cost prohibitive. Therefore, production systems have been one of the prime applications of discrete event simulation. An important consideration is the use of a statistically designed experiment so that maximum insight is gained from the system. Over the last 20 years, there has been much debate in the literature comparing the performance of Taguchi approach to design of experiments (DOE) as compared to the Traditional approach. Much of this debate has taken place within the context of product quality improvement. The Taguchi methodology has been used extensively within the domain of product quality improvement. However, the use of the Taguchi approach outside the quality domain has been relatively unexplored. Recently, it has been suggested that the task of designing production operations can be aided by the use of a Taguchi based RSM approach applied within a simulation model of manufacturing operations. The purpose of this paper is two fold. The first objective is to explore some of the issues involved when comparing the Traditional and Taguchi DOE approaches. The second objective is to highlight differences between these two approaches as supporting experimental designs for the development of response surfaces. This comparison is explored through the use of a case study from the automotive industry.