The FMEA Mini-Series: Using an FMEA vs. SxO for Prioritizing

Laura,

Your statement:

MHO? - the occurrance criteria being used should be established when you start out.....

You have it in the proverbial nutshell!!!!

Thought Process!!!!!!!!!!!!!

ASD...

Originally posted by FMEA best:

yOU CAN NOT CHANGE THE SEVERITY OF DYING ONLY THE OCCURANCE OF IT HAPPENING OR THE DETECTION OF THIS THE WARNING SHUTE IS A DETECTION METHOD SO CHANGES THE DETECTION SCORE AND NOT THE SEVERITY. tHE SMART CHUTE IS WHERE YOU ARE REDUCING THE OCCURANCE OF SOMEONE DYING SO AGAIN YOU CAN NOT CHANGE THE SEVERITY OF SOMEONE DYING BUT YOU CAN REDUCE THE OCCURRANCE OR INCREASE THE DETECTION BEFORE JUMPING.

Click to expand...

Now you're just being silly. We're getting the same verbal gymnastics again.

Just because you can cite a case where you believe the severity cannot be modified does not extend to imply, much less prove, that in no case is this possible.

To those of you who still feel that you can never modify a severity rating through a design change, I suggest you write a paper and contact the ASQC and other related organizations. Every FMEA book I have read in the last 15+ years states this basic premise. I believe if you can prove your theory you will be the first to do so and will probably go down in 'quality history'. Your revelation will be a turning point and will profit you greatly. You will cause authors to rewrite their books and training companies will change their courses. As an example, you will force the AIAG to change their statements on pages 21 and 43 of their FMEA manual.

I am anxiously awaiting such a paper as well as agreement from your peers that your premise is correct. Until then, I personally will continue as I have for over 15 years - I will modify the severity of an FMEA line item through a design change.

yOU CAN NOT CHANGE THE SEVERITY OF DYING ONLY THE OCCURANCE OF IT HAPPENING OR THE DETECTION OF THIS THE WARNING SHUTE IS A DETECTION METHOD SO CHANGES THE DETECTION SCORE AND NOT THE SEVERITY. tHE SMART CHUTE IS WHERE YOU ARE REDUCING THE OCCURANCE OF SOMEONE DYING SO AGAIN YOU CAN NOT CHANGE THE SEVERITY OF SOMEONE DYING BUT YOU CAN REDUCE THE OCCURRANCE OR INCREASE THE DETECTION BEFORE JUMPING.

I agree Marc. I am sure we could come up with a lot of examples on real FMEAs where a design change HAS reduced the severity rating.

The topic was fun for a while, but as you say don't beat the dead horse here. Take it to somebody who cares.

Dave

It has been fun and, I believe, informative and thought provoking. And I remember years ago I had trouble with the concept until I had been on several FMEA teams. I admit I was a bit 'rough' in my last post, but -- geezzzzz. I also wanted to point out that this is not just my opinion.

Halo, I am new here in this interesting forum.
In column of 'detection' , if I want to rank based on the ability to detect the cause of the failure, how can I do that. The table in FMEA manual seems only for ranking based on the ability to detect the failure. Can I mix the two approach (rank based on detection the cause and detection the failure)? the rank will be very different.
Can anybody explain to me about this? thank you.

Dear Ibrahm,

Please see the discussion in the FMEA manual.

D ranking relates to the ability of your detection system, it doesn't matter if it to detect the cause or the failure mode.

So to my understanding you can use the same ranking table for the cause as well as the failure.

Good Luck
MSAFAI

Originally posted by MSAFAI:
Dear Ibrahm,

Please see the discussion in the FMEA manual.

D ranking relates to the ability of your detection system, it doesn't matter if it to detect the cause or the failure mode.

So to my understanding you can use the same ranking table for the cause as well as the failure.

Good Luck
MSAFAI

Click to expand...

Be careful, the detection rating assumes the failure has already occurred.

Detection ratings apply to how well your current process controls will detect the failure.

I can't see where the Severity, Occurance, and Detection ratings can be interchanged.

MHO

ASD...

Dear Al,

I think there was a misunderstanding.

I was only talking about D-ranking table. I meant it can refer to the ability of a control method to detect a 'failure mode' or a 'cause'. (type 2 and 3 controls)

Regards,
MSAFAI

For MSAFAI and All Dyer, thank you for the reply. Maybe I need to explain my case more clearly. There is one failure mode, just say the wire color is red (I work in cable industry), the potential cause is too high temperature of annealing. S = 8, O = 6.
There are two Current process control in practice: check temperature 1 time per shift, and inspect 100 % of the product.
To find D:
If I use the first current control, D will be 5, so RPN = 240. Need action (I use RPN 100 as a limit)
If I use the second one, D will be 2, so RPN = 96. No need action.
What should I choose? In my opinion, I should lead my process to be failure preventive. So 100 % product inspection should be avoided. But according to the manual,that states that we can use type 2 of 3, It seems that we can choose the number of RPN. We want to high, use the second approach (in my case above). We want to low, use the first one. Am I correct?