Beyond PDCA - A New Process Management

Abstract: The ISO 9001 quality management standard calls for the use of the plan-do-check-act (PDCA) model for managing processes. The author questions why check is included in the cycle when the goal is to reduce the need for verification activities. Current demands for qu…

Access this article



Remember Me
Log-In Help
Other ways to access this article

Social Bookmarking

Digg, delicious, NewsVine, Furl, Google, Spurl, BlinkList, Simpy, StumbleUpon, BlogMarks, Facebook

I am trying to identify a textbook to utilize for PDCA training. Can you recommend one? This will be used for training healthcare personnel.
--David M. Adams, 06-14-2011

I have long enjoyed reading Praveen's articles and books. However, I find this statement rather startling:

"Implementations of Six Sigma accentuate limitations of control charts and the PDCA model. The required defect rate for a Six Sigma level process performance is 3.4 ppm, which is difficult to achieve with Shewhart's control charts designed to produce a product with a defect rate no more than 2,700 ppm beyond 3-sigma control limits. Thus you cannot achieve Six Sigma level performance of 3.4 ppm using the current implementation of PDCA or control charts."

To me, this represents a fundamental misunderstanding of the relationship between process control and process capability. Control charts do not assess the capability of the process to produce anything to specifications; they are used to assess the raw capability of the process. Three-sigma limits are just what they are: useful dividing lines for deciding what's likely to be a shift in the process or other unusual occurrence, and what's just noise. They tell us if our process is coherent and our data are homogeneous enough to make further inferences about distribution models and use tools that depend on those models, including capability studies. Estimation of DPMO for continuous measures is highly dependent on capability studies, which provide very little information in the absence of a state of statistical control.

There are many examples of companies that have achieved levels way beyond 3.4 million defects per opportunity using nothing but Shewhart Charts and PDSA. In Wheeler and Chambers, the Tokai Rika company of Japan produced cigarette lighter sockets. The record as we see it STARTS with a process that has a Cpk of 3.31, and ENDS with a process that shows a Cpk of close to 5 (maintained over 13 months). This means that it's very unlikely that during the last 13 months of the record, Tokai Rika made a single defective lighter socket, in a run of over 5,700,000 parts. This was done using a paper XmR chart with Rule 1 of the Western Electric Zone Tests (point outside the control limits).


--Rip Stauffer, 01-06-2011

Hi Praveen, you may want to look at the STARS methodology which used PDCA as an input and has gone a long way beyond PDCA. best regards Han
--Han van Loon, 03-30-2008
Sign Up for Newsletters