B
Bill Pflanz
Now that I have provided some historical information on the subject, I still want to comment on some of Praveen's postings.
As shown in my previous posting, Shewhart did not develop what has become known as the PDCA cycle. In The New Economics, Deming indicated that the most important application of Shewhart’s theories was the management of people not control charting.
Whether you call it a need for action or defining a target, the concepts of statistical thinking include the need to produce within specification limits and does not prevent one from using PDCA/PDSA.
Since Deming was a student of Shewhart, I will again quote him on specification limits. From The New Economics:
I think someone with a real understanding of statistical thinking and the PDCA/PDSA cycle will know what needs to be done. If they do not understand then switching to PPPP will probably not make them understand either.
Bill Pflanz
Praveen: Further, Shewhart developed his OCAP (Out of control action plan) to monitor the process output, and thus developed process control. In other words, he used process control to manage product flow.
Continually improving the process, now expectation is beyond control. The new expectation is management to achieve virtually perfect output. This requires management, which means input must be controlled.
As shown in my previous posting, Shewhart did not develop what has become known as the PDCA cycle. In The New Economics, Deming indicated that the most important application of Shewhart’s theories was the management of people not control charting.
Praveen: In order to understand the intent of PDCA and 4P, one must understand what Shewhart was trying to do as he is the one who developed concepts of statistical thinking, and imparted intelligence into quality work. I mean intelligence through statistical thinking, and developed the predecessor to PDCA. PDCA is an aberration of the Shewhart's cycle.
Then, it is very critical to understand what is meant by quality. If one says to produce that is acceptable within specification limits, that is one definition. If one defines quality being on target, that is another definition. Virtual perfection implies close to target. There can not be many definition of saying being on target.
Whether you call it a need for action or defining a target, the concepts of statistical thinking include the need to produce within specification limits and does not prevent one from using PDCA/PDSA.
Praveen: Second, in PDCA, in contrast to its intent, normally C is used for check against defined specification limits. Consequencially, we end up focusing on building to specification limits without defining target (a single value of performance). In the 4P, C is replaced by Perfect, that means building to the defined target, thus defining target is imperative. Now, if we build to limits, our distributions are centered around limits, thus significant variation outside the specification limits, which we try to manage through excessive inspection. When we establish targets, our distributions are closer to targets instead of limits, thus a significant reduction in variation outside the specificaiton limits. Hence, one can see defective PPM's much lower, and better Cpk's. Then, one can live without inspection, and reduced cost.
Since Deming was a student of Shewhart, I will again quote him on specification limits. From The New Economics:
Deming:Specification limits are not control limits…. A process may be in statistical control yet turn out to be 10 percent defective – 10 out of 100 items outside specifications. In fact, a process could be in statistical control yet turn out to be 100 percent defective.
A point outside specifications indicates need for action on an item, such as inspection, to try to separate good from bad. A point outside control limits indicates need for identification of a special cause, and if it can recur, removal thereof. My point is that there is no logical connexion between control limits and specifications.
I think someone with a real understanding of statistical thinking and the PDCA/PDSA cycle will know what needs to be done. If they do not understand then switching to PPPP will probably not make them understand either.
Bill Pflanz