Struggling with a root cause analysis - Customer Returns - Escape issue

[FloridaST]

Recently we received back some parts where a component was not assembled correctly. The criteria is clearly shown on the assembly drawing and our work instructions also clearly call out the requirements. This has happened one time before (I wasn't here at the time) so there is likely a systemic breakdown which is allowing this to happen. The escape issue I can fix. Where I'm struggling with is explaining WHY this happened. Why did the production staff build it improperly when I know the answer is because they weren't paying attention or just didn't care? We have very clear work instructions which clearly state the requirements. These are not new employees that build these and the parts are good parts and they still get built wrong. I think it's because they don't know why the requirement is what it is so they don't put much thought into it.

I spoke with the customer about the design of this product and their explanation made a lot of sense. I can flow this down to the staff for awareness to help motivate them to build these properly. That said, the requirement is known and they shouldn't need to know why.

Anyway, it's frustrating, maybe someone can give me another perspective.

 

[Sidney Vianna]

That said, the requirement is known and they shouldn't need to know why.

Anyway, it's frustrating, maybe someone can give me another perspective.

ISO 9001:2015 7.3.d) is applicable here. When people become aware of the (potential) consequences for sloppiness, they tend to care, especially when serious consequences might exist. Further, you can direct the root cause investigation for lack of robust verification of the assembly process. Not an elegant solution, but, if there is systemic lack of discipline, you might have to strengthen the verification of the assembly process, before product ships. Much easier to address than "lack of discipline".

So, the root cause of the escape would be lack of proper verification post assembly. If the customers can detect the problem, so should you.

Good luck.

 

[blackholequasar]

I struggle with this too. When we have a process in place, and we have a work instruction and visual aid -- and yet it STILL fails not only to be created correctly but also makes it through the quality check... it's such a process breakdown that there has to be a better way to prevent it. It comes down to if people are paying attention or not sometimes, and that variable is hard to control. You reduce the risk of someone being careless when you create a quality stop along the way, but when that fails it can feel like the process is trash. You can't root cause 'people don't care' haha

You have to believe that if there is a breakdown in the process, it can be repaired. The solution, however, isn't always super visible. So I really can empathize with your situation.

 

[Golfman25]

I would argue with your premise -- that you have a systematic issue. If this is only the 2nd time it happened (the first being when you weren't even at the company), I would ask how many units ago, or how long ago. There are some things that the cost to make perfect are huge compared to the cost of a singular return every now and then. So do that analysis to determine if you really have a "problem." Then you can work from there. Good luck.

 

[FloridaST]

Thanks Sidney. The quality escape is something that I've fixed (need to build more product to be sure) but at least a control is in place that wasn't before.

 

[FloridaST]

I would argue with your premise -- that you have a systematic issue. If this is only the 2nd time it happened (the first being when you weren't even at the company), I would ask how many units ago, or how long ago. There are some things that the cost to make perfect are huge compared to the cost of a singular return every now and then. So do that analysis to determine if you really have a "problem." Then you can work from there. Good luck.

I hear ya and to clarify, the quantities are rather large, 1000+ pieces each, and these RMAs are only about a few months apart so it's a problem. If it were only 1 or 2% ,that's one thing, but entire lot rejects need some attention.

 

[FloridaST]

Thanks for the input everybody. I did some further research on the cove and this is a fairly common topic but your responses to my personal issue has helped get me thinking differently about it. To me, there's a condition that is allowing the operators to build these improperly and that's what I'm looking to fix. Hopefully I can implement a control poka-yoke to remove some of the burden on them.

 

[Sidney Vianna]

I would argue with your premise -- that you have a systematic issue. If this is only the 2nd time it happened (the first being when you weren't even at the company), I would ask how many units ago, or how long ago.

Frequency is important, but so is the criticality of the failure. An isolated discrepancy which could lead to very serious consequences obviously should be treated via RCA and true corrective action.

 

[optomist1]

could be I missed something, but generally, when it doubt poke yoke the process or operation(s)...this may be a bit off base, but depending on the type of device and process, when sporadic issues are encountered, "witness marks" are a simple but effective method for work station pesonnel or auditors to confirm it was assembled/oriented as required. The witness marks are frequently used as part of the "safe launch" phase, in automotive field...

 

[FloridaST]

Frequency is important, but so is the criticality of the failure. An isolated discrepancy which could lead to very serious consequences obviously should be treated via RCA and true corrective action.

Right, so my example is a real actual issue. If the parts aren't aligned properly, they won't see the pin that fits between them and then an alarm will sound on the finished unit. This pin was reduced in size recently for safety reasons so now it's extremely important that the parts are aligned in the piece that we build. Before this occurred, we could get away with alignment issues but not any longer. I never would have known about the pin until I asked them how this assembly is used by them. At least now I have something to work with when explaining why it's important to our assemblers.