Design Of Experiments

Written by Nicholas Kamuda
Bookmark and Share

One of the central processes used in the application of Six Sigma, called Design of Experiments, helps Six Sigma practitioners determine the relationship between an input of a process and the outputs of that process. Also called DoE, design of experiments is an important part of the Analysis stage of the Six Sigma methodology.
Commonly, analysis is the third of five incremental stages of process improvement.

Referred to as DMAIC, the five stages of Six Sigma implementation are to Define, Measure, Analyze, Improve, and Control. Bracketing those core five are two ancillary stages, Recognize and Realize. Recognition begins the Definition process in which Six Sigma practitioners determine every possible cause that can lead to defective output. During the next stage, Measure, people work to actively collect data involving the causes and the defects, and also weigh that data against the desired outcome in order to determine what areas require the most attention.

Using Design of Experiments in DMAIC

Next is the Analyze stage, the stage in which effective Six Sigma design of experiments is critical. DoE can be used to conduct various tests on processes, determining which factors are critical in controlling the quality of output. By utilizing successful design of experiments tests, those involved can help narrow the focus from a large number of possible factors to a few key factors.

The procedures behind using a DoE are largely controlled by mathematically systematic tables that are developed by the leading Six Sigma practitioner involved (often a Black Belt). The methods are designed to induce changes in the output by subtly altering various inputs, one at a time. By carefully proceeding under this methodology, Black Belts can systematically eliminate irrelevant factors, backed by hard data.


Bookmark and Share