I recently performed an in-house product audit, and made the following discovery:

Control Plan and Inspection Record both require that a GO/NOGO gauge be used for inspection. GO member = 9.20mm, NOGO member = 9.31mm

An audit of the gauge showed that the GO member was 9.21mm; the original GO member had been lost some time ago.

When I raised this with the Quality Engineer for the program (my peer), his view on this was that since the gauge was within the tolerance band specified for the inspection, that this was not a nonconformance, but rather an OFI (opportunity for improvement).

I have been unable to find anything specific in either QS or TS that deals with something like this. I would like to solicit the opinions of any of the members here, especially those who are RAB-registered auditors (that will definitely lend weight to whichever side of this decision you lean towards).

I don't mind (and hope that I am) being told that no, this is NOT a nonconformance (less paperwork and followup for me), but I don't want to leave a gaping hole in our system that the REAL auditors will find in a few months.

Since I am unlikely to get back on tonite before going to work tomorrow morning, a brief email to the following address (ronr(at)pr-cn(dot)com) would be of great assistance in resolving this as quickly as possible.

Thanks guys

I am an IATF accredited 3rd party TS auditor if that will suffice?

It is an interesting issue. I do not see a failure or NC against the inspection process itself. If a certain range is conforming, a tighter version of that gage would still be conforming. And, since it was lost, at least they diligently tried to improve the method, rather than loosen it. (Driving 75 in a 50 mile per hour zone is not legal, but 45 in a 50 is legal).

It could clearly be at least an opportunity for improvement because conceivably, good product is probably being scrapped at that lower range.

However, if the auditor feels the responsible process manager was being sloppy or uncaring about the proper methods for handling the loss of a gage, you could write that aspect up against some of the management and oversight processes, if you feel it will take that to get him to follow his management responsibilities. But, I would only go that route if I felt it was necessary to not let him off the hook.

We run into this fairly often. Normally I handle it by adding "in-process" or "narrow limit" to both the instruction and the CP (if we're going to leave the gage as it was found). This covers not being at the print specification.

As far as the nonconformance, all I can say is that we have written up this type of issue. Yes it's more paperwork and follow-up but we also have the ability to track the issue and see just how widespread it really is (we have around 3500 gages in our system). Because of that we have improved our system somewhat when gages get "lost", damaged, or whatever. One of the improvements is that the Quality Engineer must be notified so the instructions and CP can be updated (if necessary).

Hi Ron, I've taken and passed the RAB ISO 9001 auditor course - haven't paid to register yet but still have time left to do so.

As was noted, I'd be slightly concerned that you're rejecting good product. I'd be more concerned about the loss of the 9.20 mm go gage - a good calibration system should have caught it's loss and raised a flag about product inspection. I'd also be concerned about the method used for choosing the replacement gauge - any temporary work instructions issued to reflect its use? Why wasn't another 9.20 gage purchased and brought into the system? I'd also be concerned if this is a single occurence, or are there other instances where something similar has occured?

Personally, I'd probably write it as a nonconformance, even though that it would impact me more as a QA engineer than just an opportunity for improvement. But then I've been accused of being a bit of a masochist, as well as being a perfectionist :)

I recently performed an in-house product audit, and made the following discovery:

Control Plan and Inspection Record both require that a GO/NOGO gauge be used for inspection. GO member = 9.20mm, NOGO member = 9.31mm

An audit of the gauge showed that the GO member was 9.21mm; the original GO member had been lost some time ago.

When I raised this with the Quality Engineer for the program (my peer), his view on this was that since the gauge was within the tolerance band specified for the inspection, that this was not a nonconformance, but rather an OFI (opportunity for improvement).


A few thoughts:

It's best to construct your documentation so that it doesn't make people's jobs more difficult. In this instance, you have a QE who made a judgement call in the absence of the 9.20 pin. We should be able to assume that he made his decision based on what was known about process output, for example. If the process mean was running at or near nominal and the process variation was comfortably within the specification limits, the use of a 9.21 pin isn't much more than a formality, and shouldn't be a source of controversy. The documentation should be used to allow smart, experienced people to act smart and experienced.
You should probably have a process in place for rapid replacement of broken, worn or missing gages. When it's important to have the correct gage, backups should be readily available. When they're not available, you should have a contingency plan. Of course, new gages can't be ordered if no one reports that they're needed, but that's a different issue.
When confronted with a conundrum regarding whether or not to write something up, ask the question, "Will writing this up help people to do a better job?" If it will, then do it, and make sure that operators and process owners understand that the system is being attended to, and if the system is made better, their jobs will be easier. If you're not sure that what you're doing is helpful, don't do it. Allow some time to look at the situation more carefully, and from different angles. Remember--you're not trying to enforce the law; you're trying to determine if the present method is working or not

But then I've been accused of being a bit of a masochist, as well as being a perfectionist :)
I fully empathize :yes: :agree:

Thanks for the responses and suggestions guys.

After reviewing this with the QA Manager (my boss), I have decided to write this BOTH ways.

As an OFI, there is not a clearly documented/understood procedure to be followed when a gauge is found to be missing (as opposed to out of calibration, etc).

As a nonconformance, in that our system allowed a gauge which was not properly calibrated (no evidence to show that the 9.21mm GO member had been calibrated after being selected) to be used for inspection.

I know it's a bit nitpicky, but we are preparing for our readiness audit, and it's much better if -=I=- find it than if our external auditor or (gods forbid) a customer audit does, neh?

The documentation should be used to allow smart, experienced people to act smart and experienced.




All three points were good, but this one should be embroidered and hung on every QMS Manager's wall...:applause:

than simply grading of a audit finding as a 'non-conformance' or OFI, too. What's the QE thinking? Is a replacement pin on it's way? Are you considering another audit of the calibration system to see how many more are like this to confirm it's a localized event? I'd be considering your 'awareness' program to see what people know about making ill-informed decisions like this!

Andy

I'd be considering your 'awareness' program to see what people know about making ill-informed decisions like this!

Andy
What's the "ill-informed decision" you're referring to?

as you pointed out, Jim, that the QE knew the process capability etc., - the first point you made in your response.
My experience is that little or no thoughts are given to these situations; a locally made decision to replace the pin and not spend any time discussing with the operator etc. the importance of flagging these events - it was 'just lost' and everyone shrugs their shoulders and walks away.........a replacement is found etc. etc. Of course, if they have to do all the things you listed, maybe it's a poor assumption that they knew what to do in a situation like this...........:confused:

But then, maybe someone did do all those things and you're correct in your assumptions.............it's just not been that way in my 30+ years of experience in many roles.:(

Andy

The fact of reducing the tolerance limit of the gage not always lead to refuse good product, due to it could lead to reduce the variability of our product and have parts with lower variation spec speccially if a pre-capability study is done at first step of production to detect this.

I mean:

the Lower limits in gages -> real More capability of the part to meet the real spec the heigher..

cp= Upper Limit - Lower Limit / 6Sigma

When you define a limits of one gage, then you have to built it,

and my question is.

What is the probabiility when we create a gauge that meet the value we want? the response is nearly cero.

i.e GO member = 9.20mm, NOGO member = 9.31mm

Probably the real Go member limit onced produce the gauge will be 9.2011 but maybe within spec.

For this reason it also exist a tolerance for every value.

And the real question will be:

Is the 9.21 whithin tolerance previoulsly defined for Go member?.

therefore my study is:

9.21 whinin Gage Go tolerace -> No CAR

i.e 9.20 +-0.1

9.21 not within Gage go Tolerance -> CAR

i.e 9.20 +-0.05

To summarize:

You shoud have stablished a Tolerance limit for Gage Go member and test if this number 9.21 is ot not within?.

Other limit will be required for Gage not Go member.