The work of Genichi Taguchi has recently become a subject of great interest and promise amongst U.S. manufacturing companies. Dr. Taguchi’s philosophy contains two important elements [1] :

1. Quality loss is defined as deviation from target.

2. Quality can only be achieved economically by being designed in.

Dr. Taguchi’s approach employs orthogonal arrays and analysis of variance (ANOVA) tables in place of Taylor-expansion methods[2]. This approach appears to provide an opportunity for greater accuracy and proven convenience for practical engineering problems. This paper reports on a portion of a major effort in the Design Productivity Center to add theoretical justification to the Taguchi approach. In particular we propose a new algorithm which employs orthogonal arrays and conjugate directions in an effort to make Taguchi Methods more effective and theoretically defined.

This content is only available via PDF.
You do not currently have access to this content.