Information about PPM since giving up AQL - Historical description of requirements

[A.C. Mouchet]

I am looking for an historical description of the requirements of automotive principals concerning the PPM since the giving up of the AQL. Web sites, books, reviews, ... ASAP. Thank you very much.

 

[Marc]

I don't know any specific sites, etc., but In my memory AQL started receeding in the late 1980's as the philosophy went from inspection to 'manufacturing responsibility'. I believe the personal computer has in part allowed this as the data necessary before could be overwhelming. In addition, the Cost of Quality started becoming 'interesting' especially as companies started trying to quantify the cost of Bad quality and had some degree of succeess. I first heard ppm in the electronics world. It might have been with respect to 6 sigma. The conversations I remember included simple stuff but most was based on the idea that there is no 'Acceptable Quality Level' (acceptable number of defects) and that serious continuous improvement was necessary. As I remember 6 ppm was a goal at many places. Damn, I've got the 'old timers' disease - can't rmember much any more.

Anyone know the equivalences between AQL and ppm? (Like an AQL of 95% is X ppm.)

 

[BWoods]

I don't know of a way to directly link AQL to PPM. To me that is apples and oranges.

Will this help?

+/- 1sigma = 68.27% = 317,300 ppm outside
+/- 2sigma = 95.45% = 45,500 ppm
+/- 3sigma = 99.73% = 2,700 ppm
+/- 4sigma = 99.9937% = 63 ppm
+/- 5sigma = 99.999943% = 0.57 ppm
+/- 6sigma = 99.9999998% = 0.002 ppm

 

[Marc]

Also see:
AQL (Acceptable Quality Level) vs. PPM (Parts per Million) - Equivalences?
and
AQL (Acceptable Quality Level) vs. LTPD (Lot Tolerance Percent Defective)

 

[Sam]

Marc, Didn't Crosby advocate doing away with the AQL with his "Zero Defects" process in the late 60's?
Also the Electronics Industries Association standard for verifying compliance with quality levels in PPM is EIA-555,approved in 1988, Assessment of nonconforming levels in parts per Million (PPM) EIA-554, approved 1988,and my favorite, Guideline on the use and application of Cpk,QB6, 1991.
As for published acceptable quality levels (AQL), those are outlined in the old mil std. Daimler-Chrysler publishes an an acceptable PPM level of 50.

 

[Marc]

Sam said:

Marc, Didn't Crosby advocate doing away with the AQL with his "Zero Defects" process in the late 60's?

I think that was the deal.

The problems which come into play are reality and price vs. theories when one starts speaking 'zero defects'.

 

[Atul Khandekar]

Right Sam. The real quality level IMO is determined by PPM and not by AQL.

 

[rrramirez]

Assessment of nonconforming levels in parts per Million (PPM)

In the Dr. Wheeler´s book Understanding SPC, paragraph 6.3 there are some explanation related with Cp values and PPM nonconforming. In page 130 he concluded:..."The conversion of capability values into fractions nonconforming is an operation that has no contact with reality. It is nothing more than fantasy, and the results are illusion, if not outright delusions."

IMO this a related matter with the topic in discussion.

 

[M Greenaway]

An AQL is specifically a point on the operating characteristic curve of a sampling plan, and is related totally to probabilities incurred when operating sampling inspection to a specific plan.

PPM on the other hand is a direct measure of non-conformities produced in relation to total production.

There is no strict correlation between the two.

 

[Atul Khandekar]

Re: Assessment of nonconforming levels in parts per Million (PPM)

rrramirez said:

In the Dr. Wheeler´s book Understanding SPC, paragraph 6.3 there are some explanation related with Cp values and PPM nonconforming. In page 130 he concluded:..."The conversion of capability values into fractions nonconforing is an operation thta has no contact with reality. It is nothing more than fantasy, and the results are illusion, if not outright delusions."
IMO this a related matter with the topic in discussion.

The PPM statistic thus calculated represents the 'potential' of that process to produce nonconforming product. This may have some bearing on the downstream processes, if not on what is finally shipped to the customer. However, from the point of view of the operator, PPM may not mean anything, because he wants to know where the process stands 'at this point of time'.