Subsequent failure can also be a root cause?

qualprod

Trusted Information Resource
Hi everybody

I had posted similar problem , but didnt catch well the responses,
Here is.......
Is a defective product.
At doing the investigation, the problem started when a machine had a failure, then the piece finished, then the next step, (quality guy)
failed at releasing the first piece, (was bad, and was approved), the product continued until Shipping detected the problem.

I see three causes
1 A machine failure happened
2 Operator failed at verify quick check
3 Quality guy failed

For initial root cause, we found, poor machine maintenance.

Question
At implementing action plans
Is it ok, just to manage the first root cause and implementing actions
and for the rest also take actions, but not mentioning other two root causes?
Or Can I have three root causes?

Because
If the first step hadn't failed, the others wouldn't either.
How to manage this issue?
Please help

Thanks
 

Miner

Forum Moderator
Leader
Admin
Root cause analysis is often not 100% sequential as the 5 Why process would have you believe. Most company use an enhanced version of 5 Why that looks for 3 root causes:
  1. The Defect root cause (in your cause machine maintenance)
  2. The Escape root cause (in your case, you have two branches: operator and quality
  3. The Systemic root cause. What in your QMS allowed either or both of the above to occur?
While the approach above is an improvement over a single 5 why, you sometimes have multiple root causes under one or more of the above.
 

John Predmore

Trusted Information Resource
The Attachments list has a few examples of the 3-Legged-5-Why , which is a descriptive name for what @Miner described.

Any step in the 5-Why causal sequence is an potential opportunity to prevent future recurrences. With 3 legs, you have roughly three times as many points to potentially prevent, detect, or foresee recurrences (and then take action).
 
Last edited:

Bev D

Heretical Statistician
Leader
Super Moderator
Agree with all of the above

but the big questions are why did the operator fail and why did the quality person fail to detect the defect? While these are on the escape branch they are not root causes
 

PaulJSmith

(Former) Quality Jerk
When you can see from the very beginning that you have more than one path to pursue, 5-Whys may not be the best tool. I would think a fishbone diagram would work better here.

You likely have only one root cause, but you clearly have at least one contributing factor, maybe two, depending on how you choose to pursue it.
 

qualprod

Trusted Information Resource
Agree with all of the above

but the big questions are why did the operator fail and why did the quality person fail to detect the defect? While these are on the escape branch they are not root causes
Of course ,but based on these two , is possible to get the root causes of them, e.g for quality could be lack of competency, then training process didn’t do the follow up of training, then , training guy left the company, etc. and similar analysis for operator
 

Miner

Forum Moderator
Leader
Admin
To expand further on my earlier post, the vast majority of problems can be solved using the 3-Legged-5-Why, but a smaller percentage are more complicated. When you have more complicated scenarios, you can add additional legs or branch off the existing legs on the 3-Legged-5-Why, or you can use an approach called Apollo Reality Charting, which takes it to a whole new level. I will occasionally use this approach which the problem get very complicated.

I don't care for the fishbone approach because it is often recklessly applied with little data to back it up. Then people guess at the most likely cause and apply countermeasures randomly. Of course 5-whys can be similarly abused.
 

Miner

Forum Moderator
Leader
Admin
Of course ,but based on these two , is possible to get the root causes of them, e.g for quality could be lack of competency, then training process didn’t do the follow up of training, then , training guy left the company, etc. and similar analysis for operator
Yes, these are intermediate causes. You need to take it further to identify the true root causes. And don't stop at training. Why was the training inadequate, or not provided?
 

Bev D

Heretical Statistician
Leader
Super Moderator
And competency and training are tropes. (False easy whipping posts)
How do you know it wasnt an MSA or too small sample size cause?
 
Top Bottom