A Strategic Methodology for Taguchi Design of Experiments

Abstract

Design of experiments and Taguchi methods are well-established methodologies in which only statisticians are formally trained. Design of experiments (DOE) is a powerful approach for investigating the optimal combination of process parameters in respect of quality performance. Taguchi methods (TMs), on the other hand, are used for maximizing product and process robustness by reducing variation due to undesirable external disturbances which cannot be controlled during actual production conditions. Over the past decade, both DOE and TMs have had unprecedented success in showing how statistical methods assist organizations in manufacturing high-quality products at low costs. However, recent research has shown that the application of such statistical methods by the engineering fraternity in manufacturing companies is limited due to lack of skills required in manufacturing and inadequate statistical knowledge for problem solving. In order to bridge this gap, this chapter presents a practical and strategic methodology for Taguchi methods, to tackle and solve manufacturing process quality problems in manufacturing companies. The methodology is easy to understand and readily accessible to the engineering fraternity for solving quality problems in real-life situations. The objective of this step-by-step approach is to assist industrial engineers with limited statistical knowledge to tackle process quality problems using TMs.