We have a vendor who is 17025 accredited within the scope of the services they provide. They have (and still do) provide us a good job.

For various reasons we have 'relaxed' reporting tolerances. So everything else aside, item A has a reporting tolerance of +/-1C. Item A mfg. specification/industry tolerance is always tighter than our reporting tolerance.

Now, to be clear, nothing is relaxed about the calibration procedure, reported uncertainty (however that is being done), and the overall process.

The issue: Our vendor is stating they cannot perform calibration work under the 17025 due to the relaxed tolerances. As explained to me, they have written things in such a way that all the work is to be done to mfg. tolerances and/or industry specifications. I'm saying the process, approach, adjustment criteria, etc. does not change. We just want a different failure reporting value. I'm saying if they wrote their process that way, they need to change it.

Of course... in the end... I am the customer, and have reasons for my madness!:D So there will need to be some give and take from both.

Any thoughts/ suggestions?

Brad:
Let them do their process and report based on their criteria. You are in a position to review and decide on the impact of the results. You're asking them to step outside their normal (client) requirements, so I feel it's better not to ask them to bend too much. Or pay them to do it!
:2cents:

Brad:
Let them do their process and report based on their criteria. You are in a position to review and decide on the impact of the results. You're asking them to step outside their normal (client) requirements, so I feel it's better not to ask them to bend too much. Or pay them to do it!
:2cents:

Brad,

The calbration labs always provide the error found while calibrating. If they pass a judgement on the device being 'calibrated' or otherwise, they quote the reference standard on which the decision is based.

If your acceptance criteria are relaxed, you can always define them in your system and override the calibrator's recommendation.

A client of mine has a calibtation system where he uses a device in his machine shop till it remains within acceptance limits of error and when the limit is crossed, he puts a red marker on it and deploys it in his foundry shop in the same premises. The device is no longer usable in the machine shop but is very much accepted in the foundry which has 'relaxed' acceptance criteria. Everything is well defined in my client's system and is running smoothly for over a decade now. He doesn't bother his calibrator about it.

Give it a thought. The problem can very easily be solved at your end. No need to :mad::frust: with your supplier. He most likely would not listen to you anyway!

Thank you for your prompt responses. Just a little more information in case it would influence your feelings/opinions.

This vendor does a lot of work for us. Most all of the work is conducted at our facility, with our standards, and our documentation. We have procedures to cover the work being done. Now, we currently don't have a requirement for 17025, nor do we specify that the work has to be done to it. We would like for it to be, though.

As you know, my learning curve is currently on a steep, upward daily climb with 17025:D. So I am currently trying to grasp the interpretation of the standard, my vendors interpretation of it, my own interpretation, and how they work together. While I can procedurally handle from my end, we would prefer to not have documentation with a fail on it anywhere, and then some writeup saying "Oh, it's not really a fail because....". We can do that, but it would be cleaner to revise the reporting values.

Now, the methodology and class standards does not change; only the reporting tolerance.

Not sure if this provides any information for a different opinion from you.

It would seem to me that your vendor is trying to act like he knows better than you (the customer). As someone who runs a cal lab, I have a few thoughts.

Is your vendor giving you ISO 17025 Accredited certs using your standards? He does not have control over the standards, so how can he say that they are being cared for in the same manner as they would be with his quality system. Your vendor does not seem to have a problem bending his policy to include the use of your standards. Why can't he bend to use your tolerances and accept/reject criteria?


If there is one thing that ISO 17025 is supposed to do, it is to put the ultimate responsibility on the cusatomer The customer can choose tolerances, intervals, accept/reject criteria, etc...

A previous post suggested that you should not anger your supplier because they will nt listen anyway. Who is the customer again? Oh yeah, you! Remeber one of the most important things in business. I tell myself this daily: "If I do not take care of my customer, someone else will!!!!!"



It should not be that difficult for your vendor to put a simple line on his certificate to indicate that the tolerances specified are customer specifications.


Any calibration lab who hs the nerver to tell a customer to that they will not use their specifications is being ridiculous. Find another lab who will do what you require.

Just my opinion.

This vendor does a lot of work for us.

The issue is what part of his wok is comprised by you.

Most all of the work is conducted at our facility, with our standards, and our documentation. We have procedures to cover the work being done.

I wouldn't think he does the work to your standards and your documentation. His standards would be quoted on his report and his report would be on his letterhead - not in your format. I would be happy to be wrong on this.

Any calibration lab who hs the nerver to tell a customer to that they will not use their specifications is being ridiculous. Find another lab who will do what you require.

Again the practical aspect. In these areas we are not so lucky to have multiple choices that we can boot a vendor out and bring another in just because he does not listen to me as a customer. He has his constraints, I have mine - we look for a win-win solution.

By the way, the calibrator is very well justified in issuing a certificate without a decision. 17025 does not bar him from it. As a customer you can demand it and make your own decisions on usability.

I'm a lab manager as well- 17025 Accredited.

It's a Catch 22 situation. My opinion of 17025 is "document document document". We have customers that want special tolerances, and we honor that as long as we have something from them documenting the change. It is up to us to report our findings, not to tell the customer what to do with those findings. You wouldn't believe the number of gages we get back the following year after we rejected them previously.

On the other hand, if we get an auditor in who had someone whizz in their wheaties that day, we could get in trouble. It could be a minor finding at the least- and who wants to jump through those hoops?

It sounds to me like the Quality person at your vendor is being a little overmilitant and may have been spooked by a previous finding or audit. It took us a while to overcome the pickiness and roll with the punches.

I'm a lab manager as well- 17025 Accredited.

It's a Catch 22 situation. My opinion of 17025 is "document document document". We have customers that want special tolerances, and we honor that as long as we have something from them documenting the change. It is up to us to report our findings, not to tell the customer what to do with those findings. You wouldn't believe the number of gages we get back the following year after we rejected them previously.

On the other hand, if we get an auditor in who had someone whizz in their wheaties that day, we could get in trouble. It could be a minor finding at the least- and who wants to jump through those hoops?

It sounds to me like the Quality person at your vendor is being a little overmilitant and may have been spooked by a previous finding or audit. It took us a while to overcome the pickiness and roll with the punches.

Welcome to the Cove!:bigwave:

I think you folks are getting a little closer to my thought process.

1. To me, it is far less critical to be worrying about the reporting tolerance than the uncertainties present in the measurement system being used. Lot's of worrying about the tolerance, but not enough worrying about the process leading up to that decision.:tg:

2. To reinterate-They are completing our documentation, and we are not placing any statement regarding 17025.

3. We have another vendor (who is currently not 17025) who does just what you specified above- 1) here is our uncertainty, 2) Here is the procedure and applicable details regarding the calibration, 3) here are the measurements. If we provide them reporting tolerances, they will report against those.

I hate to bring up a worn out cliche', but I really want to develop a win-win situation. I am encouraging my vendor to contact their AB and discuss this situation and work within the different provisions from the standard that I interpret is for handling such situations.

This vendor does a lot of work for us. Most all of the work is conducted at our facility, with our standards, and our documentation. We have procedures to cover the work being done. Now, we currently don't have a requirement for 17025, nor do we specify that the work has to be done to it. We would like for it to be, though.


This sounds like my job in our customer's site, in their facilities, adopting their process, documentations and requirements. Difference is that our client is ISO 17025 accredited. This turn the situation around that we are have a need to be more strigent in our practices.

For your position, the vendor / supplier can accommodate your requirements even if they are 17025 accredited, so long you agree that they do not require the job to be done compliant to their specs as they have been accredited....
I would think a proper agreement / contract can help until you are ready in 17025 requirements... This will ease them having to face their accreditation issue.

...For your position, the vendor / supplier can accommodate your requirements even if they are 17025 accredited, so long you agree that they do not require the job to be done compliant to their specs as they have been accredited....
I would think a proper agreement / contract can help until you are ready in 17025 requirements... This will ease them having to face their accreditation issue.

You have hit on the core of my issue. Why can't the work be compliant? Is a good program not dynamic? Can they not develop methodology to still do the work with compliance? Now.. I don't know enough about AB to know if the one they use is Good/Bad/Ugly. But I would think this would be something to work out with their auditor.

This, to me, is part of what we are paying them for. I'm not asking them to come in and be robots. If they have the expertise in fulfilling the requirements (and assuming, they are useful requirements) then I would think they could help develop the process to be compliant.

Thank you for sharing your thoughts.:)

We have a vendor who is 17025 accredited within the scope of the services they provide. They have (and still do) provide us a good job.

For various reasons we have 'relaxed' reporting tolerances. So everything else aside, item A has a reporting tolerance of +/-1C. Item A mfg. specification/industry tolerance is always tighter than our reporting tolerance.

Now, to be clear, nothing is relaxed about the calibration procedure, reported uncertainty (however that is being done), and the overall process.

The issue: Our vendor is stating they cannot perform calibration work under the 17025 due to the relaxed tolerances. As explained to me, they have written things in such a way that all the work is to be done to mfg. tolerances and/or industry specifications. I'm saying the process, approach, adjustment criteria, etc. does not change. We just want a different failure reporting value. I'm saying if they wrote their process that way, they need to change it.

Of course... in the end... I am the customer, and have reasons for my madness!:D So there will need to be some give and take from both.

Any thoughts/ suggestions?



I agree with Andy's point. If the lab is accredited to 17025, they are following standardized practices, and reporting results to those standards.

I will add some replies as an ISO 9001 / TS-16949 auditor. I typically audit these records which you would show an auditor during your audits. The reports would show out of tolerance conditions for some gages. The report should also show you reviewed the failured points, and signed off on them.

The auditor should ask about those, and your proper answer should be to point out that based on tight normal standards, the gages would be out of tolerance, but your needs and product tolerances are not as rigid. In your gage list or procedure, you would show the tolerances you have set for these gages, and the auditor would verify the gage results were within those ranges. A good auditor would also quickly evaluate whether the modified ranges are approrpiate for the tolerances typical for your products.

If these line up, there should be no issues. If your acceptance ranges would not be appropriate, then of course, there would be a problem. This is a fairly routine situation, from my auditor's viewpoint, though some rookie auditors may not be aware. So, write it into your procedure, and make sure it is appropriate to your operations.

I agree with Andy's point. If the lab is accredited to 17025, they are following standardized practices, and reporting results to those standards.

I will add some replies as an ISO 9001 / TS-16949 auditor. I typically audit these records which you would show an auditor during your audits. The reports would show out of tolerance conditions for some gages. The report should also show you reviewed the failured points, and signed off on them.

The auditor should ask about those, and your proper answer should be to point out that based on tight normal standards, the gages would be out of tolerance, but your needs and product tolerances are not as rigid. In your gage list or procedure, you would show the tolerances you have set for these gages, and the auditor would verify the gage results were within those ranges. A good auditor would also quickly evaluate whether the modified ranges are approrpiate for the tolerances typical for your products.

If these line up, there should be no issues. If your acceptance ranges would not be appropriate, then of course, there would be a problem. This is a fairly routine situation, from my auditor's viewpoint, though some rookie auditors may not be aware. So, write it into your procedure, and make sure it is appropriate to your operations.

Good advice, but beware: I was undergoing a QS-9000 registration audit about 10 years ago. The company was doing molded interior trim components for GM (among othe customers). Some of these components had foam-backed vinyl glued and vacuum-formed to them. The GM test for the glue adhesion involved pulling on the vinyl, with the expectation that there would be no adhesion failure at xNm of pulling force, or there would be cohesion failure of the foam (meaning the foam would come apart before the glue would fail--a good thing). The GM standard cited "cohesion failure" as one of the acceptance criteria.

Unfortunately, the auditor latched onto the word "failure" like a bulldog on a pork chop, and it was a full 45 minutes before I could convince him that "failure" in that context was a good thing. I patiently explained the process to him, showed him the language in the GM standard, demonstrated that my test report (and results) met the requirements to the letter, and even demonstrated cohesion failure by asking him to pull the label off of a bottle of Windex (the paper came apart before the glue failed--a perfect example). I finally got him to back off, but he was still pretty sure that any mention of failure in a test report was just wrong.

Good advice, but beware: I was undergoing a QS-9000 registration audit about 10 years ago. The company was doing molded interior trim components for GM (among othe customers). Some of these components had foam-backed vinyl glued and vacuum-formed to them. The GM test for the glue adhesion involved pulling on the vinyl, with the expectation that there would be no adhesion failure at xNm of pulling force, or there would be cohesion failure of the foam (meaning the foam would come apart before the glue would fail--a good thing). The GM standard cited "cohesion failure" as one of the acceptance criteria.

Unfortunately, the auditor latched onto the word "failure" like a bulldog on a pork chop, and it was a full 45 minutes before I could convince him that "failure" in that context was a good thing. I patiently explained the process to him, showed him the language in the GM standard, demonstrated that my test report (and results) met the requirements to the letter, and even demonstrated cohesion failure by asking him to pull the label off of a bottle of Windex (the paper came apart before the glue failed--a perfect example). I finally got him to back off, but he was still pretty sure that any mention of failure in a test report was just wrong.

Good point. As we have discussed here often, rookie auditors can be a problem. That is why we need to understand the requirements, and understand how our system addresses them. And patiently explain what is and is not required to an auditor. It would be nice if that were not a factor, but every field has some rookies sometimes. Even I regret some of the auditor concerns and findings I argued in my earlier years. I have become much more pragmatic in recent years, and must thank these Elsmar discussions for that. A lot of good opinions expressed here have caused me to rethink and reclaibrate sometimes. I'd like to make it mandatory for rookie auditors...:D

Good point. As we have discussed here often, rookie auditors can be a problem. That is why we need to understand the requirements, and understand how our system addresses them. And patiently explain what is and is not required to an auditor. It would be nice if that were not a factor, but every field has some rookies sometimes. Even I regret some of the auditor concerns and findings I argued in my earlier years. I have become much more pragmatic in recent years, and must thank these Elsmar discussions for that. A lot of good opinions expressed here have caused me to rethink and reclaibrate sometimes. I'd like to make it mandatory for rookie auditors...:D

Helmut, you have expressed sentiments that many of us share. We all progress and learn here; the knowledge is the reason for participation here (well... besides the paycheck:tg:) Every now and then, I go back and look at some of my old posts. I can see my learning curve here-thank you fellow Covers.:yes:

As to your concern, it is appreciated and well-founded. Our current system is fairly mature and the tolerances have met scrutiny. Although, each new auditor is a new day!:D

I just would like to think the vendor has enough flexibility within 17025 to develop alternative procedures (compliant ones) and still maintain good ground with their auditor.

Guys.....

One thing to remember.....

17025 does in fact allow the lab to deviate from published procedures IF.....

The customer requests it, which must be documented in detail to include the specific requirements.....

And the lab validates the specific changes.....if the procedure is published and accepted (e.g., NA17-20series cal procedures, ASTM methods) then the only thing to validate is the new reading/tolerance requirements, and any similar changes.

There are two responsibilities.....

The lab has the responsibility to ask and to document and validate when necessary.....

The customer has the responsibility to communicate with the lab to clearly let the lab know what is needed by the customer.

This should be a simple process when all works the way it is supposed to.

Guys.....

One thing to remember.....

17025 does in fact allow the lab to deviate from published procedures IF.....

This should be a simple process when all works the way it is supposed to.

From wide ISO 9001 auditing experience we know one thing that organisations often end up binding themselves in knots by writing extremely rigid procedures in their QMS even when the standard is highly flexible. This happens thanks to Company policy, Consultants and of course the Auditors.

It is highly likely that your lab is also in a similar situation. If they are, why cant their systems change for better to suit the customer? The practical aspect again - what is the drive for doing it? The $$ factor. How big a customer are you?

If it comes to a trade off between satisfying the management, the auditors, you, and other customers - who carries more weight? If Brad was Ford, we would not be discussing this issue here.

From wide ISO 9001 auditing experience we know one thing that organisations often end up binding themselves in knots by writing extremely rigid procedures in their QMS even when the standard is highly flexible. This happens thanks to Company policy, Consultants and of course the Auditors.

It is highly likely that your lab is also in a similar situation. If they are, why cant their systems change for better to suit the customer? The practical aspect again - what is the drive for doing it? The $$ factor. How big a customer are you?

If it comes to a trade off between satisfying the management, the auditors, you, and other customers - who carries more weight? If Brad was Ford, we would not be discussing this issue here.

Well stated, Potdar.

The customer is in fact addressing the issue within their system, probably realizing some of the wisdom in your post. While I am pretty sure we're a significant part of business for them, I would only be asking for this if 1) it's important to us, and 2) I think that it would provide benefit for the customer. If they are to grow and increase their business, they will come across other scenarios.

Thank you to everyone contributing. I'll let you know how things go.