Featured Video
This Week in Quality Digest Live
Quality Insider Features
Ryan E. Day
One more from the ‘just because we can’ files
M. Berk Talay
The automaker has been handsomely rewarded for its decades of investment in the best-selling vehicle
Andrew Maynard
Underneath the hype and the spin, some truly amazing science and engineering are taking place
Jon Speer
Avoid jeopardizing the health of your QMS
Capture 3D
GE Appliances uses ATOS 3D scanners, TRITOP photogrammetry, and ScanBox to solve engineering challenges

More Features

Quality Insider News
User-friendly database offers 3,000 classes available in 35 cities in North America
Find reputable, accredited registrars within 72 hours
Marposs Mida Laser 75P Hybrid combines a noncontact laser and touch probe in one system
At annual entrepreneurship competition, eight teams pitched business ideas, and three took home cash prizes
LEXT OLS5000 3D laser confocal scanning microscope received a Silver award
Neuroscientists train a deep neural network to analyze speech and music
American National Standards Institute (ANSI) encourages relevant stakeholders to get involved

More News

Laurel Thoennes @ Quality Digest

Quality Insider

QualiPedia: DMAIC

Sustaining quality improvements

Published: Monday, June 22, 2009 - 12:03

Define, measure, analyze, improve, control (DMAIC), developed by W. Edwards Deming during the 1950s, is a statistical and analytical method used to reduce defects by finding their root causes, eliminating them, and sustaining the subsequent improvement level.

The roots of DMAIC come from the plan-do-check-act (PDCA) cycle, a method for learning and improvement, also referred to as the “Shewhart Cycle,” developed by Walter Shewhart, the statistician who developed statistical process control (SPC) while employed at Bell Laboratories during the 1930s.

Deming successfully applied the concept of PDCA to the management system processes of industrialized organizations during the 1950s, and PDCA became known as the “Deming Wheel.” Deming developed DMAIC to guide quality projects of existing business processes in a continuous effort to reduce defects.

Six Sigma is based on the continuous cycle of identifying and eliminating all sources of variation in process performance that significantly influence cost and quality. Therefore DMAIC was quickly recognized as a means of achieving the goals of Six Sigma projects. So harmonious are the five steps of DMAIC with project goals, it is now one of the two key methods on which Six Sigma projects are based.

In Six Sigma, DMAIC is basically a system for improving existing processes that perform or provide output below specifications by identifying solutions based on the collection and analysis of relevant data, and sustaining the improved process by measuring and monitoring its performance.

The following information describes the five steps of DMAIC when used in Six Sigma projects:

Define—Define the problem or the opportunity for improvement. The problem should be clearly established in quantitative terms. All sources of variation, not just the current cause, must be specifically defined so that Six Sigma projects, improvement processes, and the goals relate to the true root causes of the problem.

Measure—Measure the process performance. Collect data and take a base measurement to which future measurements can be compared and used to determine the factors that have influence concerning the outcome of the process or procedure.

Analyze—Analyze the collected data. Use statistical tools to find the root causes of the problem and determine if the problem is solvable or random, and whether the process can be improved or should be redesigned.

Improve—Improve the process. Provide solutions based on data analysis and using techniques such as design of experiments.

Control—Control the improved process. Use data collection and control mechanisms so that any deviations from the target are apparent and corrected before causing defects.





The Six Sigma Fieldbook: How DuPont Successfully Implemented the Six Sigma Breakthrough Management Strategy, by Mikel Harry, Ph.D., and Don R. Linsenmann (Currency Doubleday, 2006)


About The Author

Laurel Thoennes @ Quality Digest’s picture

Laurel Thoennes @ Quality Digest

Laurel Thoennes is an editor at Quality Digest. She has worked in the media industry for 31 years at newspapers, magazines, and UC Davis—the past 22 years with Quality Digest.


DMAIC not developed by Deming

I would disagree that DMAIC was developed by Deming. Deming dveloped PDSA. In my understanding, the intent of the P and the S in PDSA is much different than the D and the A of SS, as you have defined them. For Deming, Planning was about understanding the theory or method by which the system will attain its results. Planning was about a scientific approach to management. It was not about planning only a specific experiment. It was much greater than that. Planning was about applying the scientific method to management in order to be able to predict results. Without theory there is no leanring.

Study was as much about studying the execution of the plan in the Do phase, as it was about studying the results of the Do phase. We need to know that the plan and its assumptions held, and we compare the results to the predicted results. After this, we compare the results to the intended aim of the system to see if the results support the aim. 

This is my understanding and I think the intent of PDSA is a more wholeistic, scientific approach to management and prediction.

Deming did not create DMAIC

I must agree with the previous poster @DVANPUTTEN, Deming did not create DMAIC. In addition to what he posted, MAIC was developed by Motorola. GE actually developed the Define phase. When I went through my Six Sigma Black Belt training at GE back in 1997 there was no Define phase. GE developed it to establish the business case for each project and avoid those expensive science projects.