Mistake Proofing on Control Plan

S

SteelWoman

We had our last QS audit this week prior to shifting to TS. During lunch one day our auditor pointed out the Appendix in TS manual that lists all the things that have to be part of a Control Plan - she particularly pointed out the Corrective Action and Mistake Proofing lines.

I understand well enough how to show/link Corrective Action on a Control Plan, but am not sure I understand how to add Mistake Proofing to a control plan? Most of our "mistake proofing" is electronic/technology-based - ie, if you enter a specification outside of customer tolerance you get a flag about it.

Any help?
 
T

Tom W

We recently had our TS2 audit, and this very topic came up. The auditor wanted us to show him mistake proofing on the control plans. Through a very length discussion it was determined that this is a hard topic to think of in terms of "putting it on the control plan".

The auditor was looking for mistake proofing in terms of removing human intervention and improving the processes capability, elimination of potential nonconformances. Again this can include such a wide range of projects and topics that it was a struggle for the auditor to even determine compliance.

He wrote an OFI for this and then stated that with all of our technology advances and electronic controls that we have put in place that our mistake proofing is almost non-existant at this point. Although we continually look to improve the processes, we can only go so far and still be cost effective in the service we provide. This is one we have struggled with and probably will for a while. :)
 

Howard Atkins

Forum Administrator
Leader
Admin
The mistake proofing methods can be placed in the column for control method.

SteelWoman
if you enter a specification outside of customer tolerance you get a flag about it.
Spec tolerance - XXXX
Evaluation/ measurement technique - electronic sensor
Sample Size: 100%
Frequency- continuous​
Control Method - Line stops/ flag/ automatically ejected etc​

I think this does it
 
B

Bigfoot

Howard Atkins said:
The mistake proofing methods can be placed in the column for control method.


Spec tolerance - XXXX
Evaluation/ measurement technique - electronic sensor
Sample Size: 100%
Frequency- continuous​
Control Method - Line stops/ flag/ automatically ejected etc​

I think this does it

I think Howard is on the right track, excellent suggestion!!

IMHO: One thing I would add is to take the noted controls and tie them to the PFMEA showing that you have completed a review (via the CA/PA) and integrated it into the system to lower the Customer risk through the improved detection / lowered occurrance. If it doesn't require Human intervention it would show in the "Controls to Prevent" column of the PFMEA and likely be documented through a Process change notice or Eng Change of some kind.

A question for all of the Cove Rats out there: Do you use a regular system of review on your PFMEA and if you do are the changes, if any, resulting from it documented through your change control system? Just wondering?
 

howste

Thaumaturge
Trusted Information Resource
Yup, I think Howard is right on target here.

One more thing to consider. If a mistake-proofing device isn't working properly, it's probably worse than not having one at all because you are counting on it being there. I also add verification checks of mistake-proofing devices to the control plan.

Examples:

For a sensor that detects proper orientation of a part, you might run a part in the wrong position once a day just to make sure it detects it.

For a device that checks size limits, you could have a couple of samples - one at the limit (painted yellow) and one outside the limit (painted red) - that you run periodically to make sure the device is making good judgement calls at the limit.
 

Howard Atkins

Forum Administrator
Leader
Admin
howste is absoultly correct here
All devices must be checked on a reqular basis, it could be part of the set up procedure to ensure that it works otherwise it gives you a false level of confidence. nothing worse than people saying but the machune approved it with out any base for this "objective evidence"
 

bpritts

Involved - Posts
One of my client's customers (American Axle, GM's spun off axle operations)
mandates exactly the controls that howste and Howard suggest. They
refer to the process as a "rabbit", although I have never had an explanation
of the origin of the term.

Not only can these controls eventually fail (which I have seen) but a colleague
of mine found a situation where well-meaning but ignorant plant personnel
have disabled these checks when they slowed down production !(VUR(*%&!)
Seems that they were rejecting too many parts!

Getting back to the main point of the post, I would think that the CP is an
appropriate place for those error-proofs that measure, test, or otherwise
check something. But I think that the very best error proofs -- for example,
designs that can only be built the "right way" probablywould not belong
on a control plan. They should be recorded on an FMEA, tho.

Brad
 

Satellite

Involved In Discussions
bpritts said:
They
refer to the process as a "rabbit", although I have never had an explanation
of the origin of the term.

For you trivia hounds (pun intended)...I believe the term came from dog trials (hunting tests). Where the dog (your machine) is tested to point out or flush (detect) the rabbit or bird (test block). :bigwave:
 
B

ben sortin

A design review audit called out on the prototype control plan can verify and validate the identification and implemention of mistake proofing in the form of design features and practices. I have seen mature products with too many sensors at the process level that have high producer risks.

"Plan on checking them all unless of course you plan for not checking any."
 

PlayPaul

Registered
I think Howard is on the right track, excellent suggestion!!

IMHO: One thing I would add is to take the noted controls and tie them to the PFMEA showing that you have completed a review (via the CA/PA) and integrated it into the system to lower the Customer risk through the improved detection / lowered occurrance. If it doesn't require Human intervention it would show in the "Controls to Prevent" column of the PFMEA and likely be documented through a Process change notice or Eng Change of some kind.

A question for all of the Cove Rats out there: Do you use a regular system of review on your PFMEA and if you do are the changes, if any, resulting from it documented through your change control system? Just wondering?


Hello. I think this suggestion brings a confusion between Control method and reaction plan. All these actions: Line stops/ flag/ automatically ejected are clearly reaction plans or what's the reason behind this logic?
 
Top Bottom