Design of Experiments (DOE) is a technique that is used
to optimize the process of conducting experiments by determining what effect
adjusting various input variables will have on the outcome of a process or
system. Through the use of statistics, DOE can identify the input variables that
have the greatest impact on product quality and performance. By maximizing the
amount of information obtained from each iteration of testing and analysis
(experimentation), DOE reduces the amount of time and resources required to
determine the design and manufacturing processes that will ensure production of
a high quality product. Adapting DOE methodology to the production environment
is commonly referred to as the Taguchi Method. When properly applied, DOE
identifies the optimum setting for manufacturing processes, thereby improving
the quality of the final product while simultaneously reducing product
Typically, product design is conducted in two major steps
- system design and tolerance design. Using an iterative design-and-test
approach, the traditional process begins with the development and testing of all
components and related processes. Once the initial design phase is complete,
acceptable limits for product and process variation are then determined. To
streamline the design process, DOE/Taguchi Methods introduce an intermediate
step, called parameter design, that uses statistical tools, such as linear and
orthogonal arrays, to identify those factors that may adversely affect the
design and manufacturing process. Identifying and then controlling only these
factors optimizes the manufacturing process and keeps product and process
variation to a minimum. Also, the amount of time spent in the design phase is
shortened, which reduces design costs.
In the application of this technique, input variables are
classified as either controlled factors or noise factors. Controlled factors are
those variables over which the product developer has some control, such as the
selection of the manufacturing processes and material used for production. Noise
factors are uncontrollable variables or those over which the developer has
minimal control, such as the quality of the parts and material received from a
supplier. DOE can be used to identify those controllable and noise factors that
have the greatest impact on product performance and quality. This technique can
also be used to determine the optimum specification and tolerance values for
each of these factors in order to minimize their influence on product and
DOE is often an integral part of a total quality approach
to manufacturing and can be used to identify solutions to quality problems in
production. The adjustments required to alleviate the problem can be identified
with minimal testing and evaluation, thereby reducing the number of parts
produced that do not meet specifications and therefore have to be rejected
Through application of this technique, it is possible to
determine the influence of the individual factors on the overall performance of
the process or product being developed. DOE can be applied to producibility as a
means of more completely understanding a manufacturing process and its effects
on a product. It can help to identify and optimize critical process parameters,
assess the effects of processing alternatives on a product, and assist in the
solution of any manufacturing problems related to quality. Successful
application of this technique to product development, particularly design and
manufacturing, improves the consistency of product performance, which improves
product quality. DOE also minimizes the costs associated with product
development through optimization of the design and manufacturing processes.
Anderson, V. L., & McLean, R. A. (1974). Design of Experiments: A Realistic Approach. New York: Marcel Dekker.
Fowlkes, W. Y., & Creveling, C.M. (1995). Engineering Methods for Robust Product Design: Using Taguchi Methods in Technology and Product Development. Reading, MA: Addison Wesley Longman.
Hicks, C. R. (1982). Fundamental Concepts in the Design of Experiments (3rd ed.). New York: Saunders College Publishing.
Montgomery, D. C. (1991). Design and Analysis of Experiments (3rd ed.). New York: John Wiley & Sons.
Peace, G. S. (1993). Taguchi Methods: A Hands-On Approach. Reading, MA: Addison Wesley Longman.
Pukelsheim, F. (1993). Optimal Design of Experiments. New York: John Wiley & Sons.
Taguchi, G. (1986). Introduction to Quality Engineering: Designing Quality into Products and Processes. New York: Quality Resources.
Taguchi, G. (1987). System of Experimental Design: Engineering Methods to Optimize Quality and Minimize Costs. New York: Quality Resources.
Taguchi, G., & Yokoyama, Y. (1994). Taguchi Methods: Design of Experiments. Dearborn, MI: ASI Press.