Quick question about RPNs - Any RPN over 100 should be actioned?

[ac_london]

OK, I'm still so new to all this but after a reasonbably successful PPAP submission, the comment was made by the customer on our FMEA, that any RPN over 100 should be actioned :mg:

I've browsed threads here for inspiration and have come across "action all severities > 7" etc but I always thought that RPNs were more relative. ie. action the highest scores but not necessarily every point scoring over 100.

Yes? No?

 

[Rob Nix]

Welcome, ac_London

They are more relative than absolute; HOWEVER, If your CUSTOMER has a bogey established, you're stuck with it.

Severity ratings over 7 most often require action. You and your customer have to determine the criteria used in advance. There are no FMEA LAWS.

I prefer to either "paretoize" the RPNs and choose the highest ones, or simply select, during the exercise, which failure modes require action and which don't - period.

Of course (as you mentioned) there are several threads here in the Cove that examine this issue in spades.

 

[The Taz!]

Welcome from NEW England

They are more relative than absolute; HOWEVER, If your CUSTOMER has a bogey established, you're stuck with it.

100% Agreement

Severity ratings over 7 most often require action.

IMHO, Where you have design control to design out the high potential risk. Other than that, it is inherent in the product in use. I'm used to severities of 10 where there is an environmental impact such as a gas spill. . . or potential for a end-user to blow themselves up.

You have 100% control over the Detection. . .the Occurrence is a latent indicator of the effectiveness of Detection. Will not come down immediately.

You and your customer have to determine the criteria used in advance. There are no FMEA LAWS.

As early during APQP as possible.

I prefer to either "paretoize" the RPNs and choose the highest ones, or simply select, during the exercise, which failure modes require action and which don't - period.

Very good practice for continual process improvement also.

Good luck

 

[Claes Gefvenberg]

Hello, and welcome to the Cove, ac

You and your customer have to determine the criteria used in advance. There are no FMEA LAWS. I prefer to either "paretoize" the RPNs and choose the highest ones, or simply select, during the exercise, which failure modes require action and which don't - period.

Yep. That's what I do too. However, I don't often miss an opportunity to pick "low hanging fruit" either... If you spot something with a relatively low RPN that you know you can fix in just a few minutes or so, by all means go for it. Easy victories, particularly early on, can work wonders for the morale and that is a key to success.

/Claes

 

[Wes Bucey]

Welcome from NEW England

IMHO, Where you have design control to design out the high potential risk. Other than that, it is inherent in the product in use. I'm used to severities of 10 where there is an environmental impact such as a gas spill. . . or potential for a end-user to blow themselves up. Good luck

Often, it is much more useful to have a frank and open discussion with the customer regarding views about interpreting Failure Mode and Evaluation Analysis (FMEA) before you begin, but when these little snippets arise after the FMEA, such as the Risk Priority Number (RPN) issue, it helps to refresh your memory about all the aspects as generally used by industry before you start negotiating.
Here's an interesting "background" essay on the topic: https://www.reliasoft.com/newsletter/2q2003/rpns.htm

Suffice to say, I worry more about the net effect of a nonconformity than some artificially derived number such as (RPN = Severity x Occurrence x Detection.)
Just think - you can reduce the RPN by being less diligent in your detection methods! What would your customer think of that "action?"

 

[The Taz!]

Suffice to say, I worry more about the net effect of a nonconformity than some artificially derived number such as (RPN = Severity x Occurrence x Detection.)
Just think - you can reduce the RPN by being less diligent in your detection methods! What would your customer think of that "action?"

Amen to that!

 

[ac_london]

Thanks everyone. It IS as I suspected - that it more a question of judgement and common sense as opposed to "a set in concrete" rule. It's just I'd heard so many different versions, I just wanted to check.

Unfortunately, we stumbled into this situation having never supplied automotive parts before, and "oh, you need to do a PPAP for this part" was thrown in as casually as "would you like a coffee?" :mg: So here I am!

Claes, I also like actioning the easy ones. You're so right that it boosts you that bit more before tackling a right stinker.... and it fills the page out a bit more too

Anyway, thanks again to all that replied. Now, I s'pose I'd best away and put this info to use!

 

[Howard Atkins]

Customers usually expect that you have defined a number above which you act.
100 is common.
Powertrain mandate 40 if my memory serves me correctly

 

[Smitty C]

We have found that since the third edition FMEA book came out, our customers have been more receptive to receiving "Real Numbers"... We used to have to cook all the numbers to be below 100, to get part approval... Now our customers seem to prefer us to select occurence values based on data (in process and at customer), detection values as clearly outlined in the third edition and realisitic severity numbers...

Imagine that, now are FMEAs are actually useful tools for US to determine where our process needs attention - not just another document for our customers!!!

 

[The Taz!]

Shazam!