Gage R&R on Visual Inspection
- Thread starter paccnc
- Start date
re: Gage R&R on Visual Inspection
My control plan contains a packaging process where our shipping department puts the parts in boxes and labels the boxes. Our evaluation technique is visual for this process to make sure the correct parts go into the right size box with the correct label. Since this states "Visual" we were requested to do a Gage R&R on this. The other 2 processes (Final Inspection, Dock Audit) also state visual as well so they asked for Gage R&R for them as well. It just took me by surprise since I have never been asked for this in the 25 years that I have been in this position. So I thought I would ask all of you on the Cove if you have been asked to do this before and if this was the norm.
My control plan contains a packaging process where our shipping department puts the parts in boxes and labels the boxes. Our evaluation technique is visual for this process to make sure the correct parts go into the right size box with the correct label. Since this states "Visual" we were requested to do a Gage R&R on this. The other 2 processes (Final Inspection, Dock Audit) also state visual as well so they asked for Gage R&R for them as well. It just took me by surprise since I have never been asked for this in the 25 years that I have been in this position. So I thought I would ask all of you on the Cove if you have been asked to do this before and if this was the norm.
No and no.
The requirement (7.1.5.1.1) states "Statistical studies shall be conducted to analyse the variation present in the results of each type of inspection, measurement, and test equipment system identified in the control plan." If whoever asked you to do this believes that the people doing the inspection are an "inspection equipment system" or a "gage" then they would be correct. I don't think so.
Personnel are addressed in clause 7.2. For personnel we ask for evidence that their competence has been evaluated. For equipment we ask for calibration traceability to international standards. For the combination of equipment and personnel we ask for measurement system studies.
Yep. That's what I thought. Wonderful, huh. How stupid. But it is one of the area of emphasis these days -- control plans and msa. You can do an attribute Gage R&R. Make your good/bad samples obvious. Get the data and send it in to the clown. Or what we did was take it off the control plan. Now we just reference "visual" inspections in a note. The whole thing is a complete waste of time.
I can understand the request. I worked at a plant that manufactured rear suspension products for automotive. There was a right hand and left hand version of everything plus SE/GT versions, and our number one customer complaint was mislabeling the different versions. Not only were the physical differences subtle, but the part numbers on the labels often differed by only one character.
But how is that solved by a gage R&R?
So we have 3 inspections: Final inspection, a packaging inspection and a dock audit.
I can see why the auditor asked for MSAs for these ‘controls’ as I will explain below. I do NOT agree that the OP or anyone should perform a ‘rigged’ test just to check the box and ‘satisfy’ the requirement. This is not only wasteful and dishonest, it sidesteps a real opportunity to assess these controls and improve - or prove - the actual effectiveness of these inspections. OR perhaps even demonstrate that they are ineffective and change them or eliminate them as waste. And isn’t that our goal as quality professionals?
From the responses I am assuming pacnc - and others - have no issues with an attribute study on the final inspection? This clearly falls in the realm of requiring a “Gage R&R” as it is a primary means of control and visual inspection is notoriously difficult to effectively implement. Understanding the repeatability of inspectors will aid in improving the effectiveness of the inspection.
The packaging inspection is a slightly different beast. As I understand it from the OP’s description, it is a ‘self check’ that the packager is packing the correct part and applying the correct label. This doesn’t fit the ‘final inspection’ model as it is not independent and isn’t looking for the traditional ‘defect on a part’. It is looking for an incorrect part and/or incorrect label. As Miner points out a ‘defect’ here is just as bad as a ‘defective part’. And it’s not uncommon. I recently dealt with a similar process where the wrong label was put on two different chemicals and thus placed in the wrong inventory locations. These chemical were then picked and provided to manufacturing who luckily determined that the wrong chemicals were used in the manufacturing of two different products. The products did not pass QC testing. We caught it but the scrap was very costly and if they had escaped to the field it would have been even worse. This kind of ‘visual’ inspection doesn’t really require a typical attribute study, it really only requires an honest assessment of the opportunity for making a mistake. And then taking the appropriate actions to PREVENT the ability to make a mistake. It is important here to remember that a “gage R&R” is only one tool for accomplishing a Measurement SYSTEM assessment. Not all information and/or knowledge can be reduced to a single ‘score’, nor should it.
As for the dock audit, I would take a different approach to the MSA. In these cases the MSA should focus more on the sample size than on the visual effectiveness. A Dock audit is usually a secondary inspection; or as I like to call it “a last desparate attempt to catch an escape”
Usually the sample size of a dock audit is what contributes most to the ineffectiveness of the audit to catch anything. Unless the defect rate is really high a small sample size woul not have much chance to catch anything. And in this case it would be better to improve the upstream final inspection or better yet the process(es) creating the defects. If you have a low defect rate escaping form your manufacturing process a small sample size is usually just a waste of time. Again this MSA is not a typical Gage R&R, but can be very informative...
I can see why the auditor asked for MSAs for these ‘controls’ as I will explain below. I do NOT agree that the OP or anyone should perform a ‘rigged’ test just to check the box and ‘satisfy’ the requirement. This is not only wasteful and dishonest, it sidesteps a real opportunity to assess these controls and improve - or prove - the actual effectiveness of these inspections. OR perhaps even demonstrate that they are ineffective and change them or eliminate them as waste. And isn’t that our goal as quality professionals?
From the responses I am assuming pacnc - and others - have no issues with an attribute study on the final inspection? This clearly falls in the realm of requiring a “Gage R&R” as it is a primary means of control and visual inspection is notoriously difficult to effectively implement. Understanding the repeatability of inspectors will aid in improving the effectiveness of the inspection.
The packaging inspection is a slightly different beast. As I understand it from the OP’s description, it is a ‘self check’ that the packager is packing the correct part and applying the correct label. This doesn’t fit the ‘final inspection’ model as it is not independent and isn’t looking for the traditional ‘defect on a part’. It is looking for an incorrect part and/or incorrect label. As Miner points out a ‘defect’ here is just as bad as a ‘defective part’. And it’s not uncommon. I recently dealt with a similar process where the wrong label was put on two different chemicals and thus placed in the wrong inventory locations. These chemical were then picked and provided to manufacturing who luckily determined that the wrong chemicals were used in the manufacturing of two different products. The products did not pass QC testing. We caught it but the scrap was very costly and if they had escaped to the field it would have been even worse. This kind of ‘visual’ inspection doesn’t really require a typical attribute study, it really only requires an honest assessment of the opportunity for making a mistake. And then taking the appropriate actions to PREVENT the ability to make a mistake. It is important here to remember that a “gage R&R” is only one tool for accomplishing a Measurement SYSTEM assessment. Not all information and/or knowledge can be reduced to a single ‘score’, nor should it.
As for the dock audit, I would take a different approach to the MSA. In these cases the MSA should focus more on the sample size than on the visual effectiveness. A Dock audit is usually a secondary inspection; or as I like to call it “a last desparate attempt to catch an escape”
Usually the sample size of a dock audit is what contributes most to the ineffectiveness of the audit to catch anything. Unless the defect rate is really high a small sample size woul not have much chance to catch anything. And in this case it would be better to improve the upstream final inspection or better yet the process(es) creating the defects. If you have a low defect rate escaping form your manufacturing process a small sample size is usually just a waste of time. Again this MSA is not a typical Gage R&R, but can be very informative...Similar threads
