Accreditation of In-House Labs to IS0-17025

[T Taylor]

I seen a short article by Michael Levy who is the practice manager of quality systems for the Virginia Philpott Manufacturing Extension Partnership and who is working in a cooperative arrangement with NIST. His article raised a question with me that I have not been able to find an answer too. In his article he stated; "that those laboratories that are tightly linked to production processes are required to seek accreditation." My understanding to date had been that if we are not selling lab services then we were not required to seek accreditation for an in-house lab. By the way we are a manufacturer of automotive parts for both OEM and the aftermarket. We do business with the big three as well as with a number of other companies that subscribe to QS9000. Can anyone shed some light on this subject? I went through several posted threads but did not find a specific answer to this question.

 

[Jerry Eldred]

I won't claim to be an expert on this. However, as I work for a company providing to the Big Three (QS9000), this is near and dear to my heart.

Last I have heard on this is pretty confusing. I'll do it in bullet format in as chronological an order as I can muster.

-QS9000 3rd Edition - internal labs don't need to be accredited, but their subcontracted cal vendor labs do.
-TS16949 (the ISO answer to QS9000). This is not required, but an option for those QS9000 companies who wish to do so (go to www.aiag.org and click around. There is usually pretty up-to-date info). For those companies who convert to TS16949 1st Edition, it has the added requirement that their internal labs become compliant to ISO17025. It does NOT say internal labs must become accredited. I would be interested in seeing any document that says internal labs need accreditation

--- BUT WAIT, THERE'S MORE...

-TS16949 2nd Edition in the works (I just learned of this. I have little info on it. Just as I am now doing my gap analysis for ISO17025 compliance, I read in an article that 2nd Edition of TS16949 is supposedly rescinding(sp?) requirement for internal labs to become accredited. Furthermore, according to the article, the requirement for external calibration labs to become ISO17025 accredited may be softened.

So the latest info I have says nothing about internal labs requiring accreditation to ISO17025. This seems to be rapidly changing, as there is (apparently) draft work on TS16949 2nd Ed., in progress. I have been asking around a lot lately for some good answers to this question. I am working on external vendor quality systems development, and not sure what to require of them until the ISO17025 landscape settles down.

I don't know that that is an answer.

 

[T Taylor]

Thanks Jerry,
I feel a whole lot better after reading your comments on the subject. I could not see where it would be value added to get ISO-17025 accreditation for our in house lab. I don't know what it would cost but it has to be thousands.

 

[Ryan Wilde]

Originally posted by T Taylor
Thanks Jerry,
I feel a whole lot better after reading your comments on the subject. I could not see where it would be value added to get ISO-17025 accreditation for our in house lab. I don't know what it would cost but it has to be thousands.

Let me preface this with the fact that I am a strong supporter of accreditation. Much of my belief in it are because of the fact that I have seen many very scary practices used in a lot of calibration labs.

That said...

Unless you have a high reject rate of parts that you manufacture, or your company is going to start making parts with very tight tolerances, then you are probably correct - accreditation is not value added.

Whether or not your uncertainty on a pair of calipers is <0.001" when your tightest tolerance is 0.030" is a moot point (and don't laugh, I had a customer with very nice electronic gaging, and their tightest tolerance was in the 0.050" range - some test equipment salesman should be drawn and quartered).

If you are making parts or assemblies that meet or exceed customer requirements for quality every time, then your calibration system seems to have the best objective evidence available as to its adequacy - satisfied customers with quality products.

Ryan

P.S. Thousands is a small estimate - the last I heard was in the range of $80-90k per discipline (dimensional, electrical, etc.) because faith is cheap, but ironclad proof is very costly. Just remember though, proof is often cheaper than rejected product...

 

[Jerry Eldred]

The only addition I would make is that there are two topics in progress above:

1. Are in-house labs at QS9000 companies going to be required to be accredited or required to be making progress toward it?

-- The answer to that original question is no.

2. Is there sense to seeking accreditation by those same labs?

-- The answer to that may by different.

I believe I understood the original question to be (1) above. Regarding (2) above, I have mixed feelings. The value of any certification or accreditation is not in the 'certificate' but in the preparation, and in having a knowledgable, objective set of eyes assess that preparation work. The value in the assessment is only that it has the tendency to drive those in the lab to a level of excellence they may not achieve without it. It also resides in the new attention to detail added during the process. I would have to say (as I have about 25 years in 3rd party labs, in-house labs, and government labs, and have started a few) that the integrity of the people in the lab no matter what, ends up being the real bottom line as to the adequacy of calibrations performed. No accreditation in the world can overcome an integrity problem.

However, in a comprehensive, honest, multi-disciplined lab environment, an objective assessment to preparations made to adhere to ISO17025 could certainly be of value. But I would stop short of making it a requirement.