Emphasizing Severity more than Occurrence and Detection vs. Overall RPN

[vanputten]

Has anyone heard about the auto industry (or specific auto manufacturers) placing emphasis on the Severity score and not the overall Risk Priority Number (severity x occurrence x detection)?

It was suggested to our organization to determine action plans based on severity only and not the overall RPN.

Thank you,

Dirk

 

[errhine]

Re: Emphasizing Severity more than Occurrence and Detection

Those in the auto industry that I know will only use RPN. Even if you have a severity of 10, if you almost always detect that failure, and it almost never occurs, it would be foolish to spend increasing amounts of money to try to reduce a severity. You would be better off to reduce your occurrences and increase detection of less severe items (given a high RPN). Thats the whole point of the FMEA.

 

[Al Rosen]

Re: Emphasizing Severity more than Occurrence and Detection

Those in the auto industry that I know will only use RPN. Even if you have a severity of 10, if you almost always detect that failure, and it almost never occurs, it would be foolish to spend increasing amounts of money to try to reduce a severity. You would be better off to reduce your occurrences and increase detection of less severe items (given a high RPN). Thats the whole point of the FMEA.

How can you reduce severity?

 

[Stijloor]

Re: Emphasizing Severity more than Occurrence and Detection

Has anyone heard about the auto industry (or specific auto manufacturers) placing emphasis on the Severity score and not the overall Risk Priority Number (severity x occurrence x detection)?

It was suggested to our organization to determine action plans based on severity only and not the overall RPN.

Thank you,

Dirk

Hello Dirk,

That's correct. There's too much emphasis on RPN. That can be very misleading. RPN is a simple multiplication of 3 numbers. Each of the numbers (rankings) has a distinct meaning. In terms of prioritizing what actions to take? One OEM (Ford) recommends to go after Severity first. See if a design change can reduce severity. Next, look at Severity x Occurrence. Finally look at remaining RPN's.

Sadly, some automotive suppliers play an RPN numbers game. They develop an FMEA procedure, state that below an RPN of 100, no further action is required. Guess what I see on their FMEA's? Exactly! No RPN above 100.

Upon PPAP submission, often no response from the customer. And so it goes. Another worthless document for a PPAP "package."

Stijloor.

 

[Stijloor]

Re: Emphasizing Severity more than Occurrence and Detection

How can you reduce severity?

Hello Al,

Through a design change; although this may not always be feasible. But trying to reduce severity is the first "order of business" when designing a product and/or process.

Stijloor.

 

[Miner]

Has anyone heard about the auto industry (or specific auto manufacturers) placing emphasis on the Severity score and not the overall Risk Priority Number (severity x occurrence x detection)?

It was suggested to our organization to determine action plans based on severity only and not the overall RPN.

Thank you,

Dirk

I have seen one prioritization scheme that had three categories:

• Priority 1 = Safety (Severity ratings of 9 or 10)
• Priority 2 = Warranty (Severity rating of 6 - 8 with an Occurrence rating > 5 and a Detection rating > 3)
• Priority 3 = Top 5 RPNs remaining after Priority 1 and 2 have been addressed

I incorporated it into our FMEA process and liked it. Never had any audit issues regarding it either.

 

[vanputten]

Thank you everyone. This is great information.

Has anyone performed FMEA's using both the customer and consumer in ranking Severity as suggested by the 4th edition of the PPAP manual?

Thank you,

Dirk

 

[someshvar]

Can anyone tell me how would you reduce severity for a said failure mode?

As per best of my knowledge even if you improve the design, you can reduce occurrence and not severity.

I am looking forward to serious discussion as to how would you reduce the severity.

Thanks and regards,
Hardik Someshvar

 

[Miner]

Can anyone tell me how would you reduce severity for a said failure mode?

As per best of my knowledge even if you improve the design, you can reduce occurrence and not severity.

I am looking forward to serious discussion as to how would you reduce the severity.

Thanks and regards,
Hardik Someshvar

You can change a design to eliminate the failure mode entirely, or by adding a backup.

An example would be a bonded metal to rubber engine mount. Most modern designs include a mechanical interlock such that if the rubber fails (e.g., tears, delaminates from metal, etc.) the metal components keep the engine from dropping.

The severity changes from engine falls to pavement to harsh vibration and noise. A clear drop in severity.

 

[Jim Wynne]

Can anyone tell me how would you reduce severity for a said failure mode?

As per best of my knowledge even if you improve the design, you can reduce occurrence and not severity.

I am looking forward to serious discussion as to how would you reduce the severity.

Thanks and regards,
Hardik Someshvar

Any design improvement that reduces risk of injury (for example) would likely reduce the severity factor. I understand the confusion; it stems from the idea that if some phenomenon x has an explicit risk of y, no matter what you do, the phenomenon as defined will have the same degree of risk. Reduction in the severity factor assumes that the problem will be redefined such that the former condition has been reasonably mitigated. Think of automobile airbags as an example. The risk of serious injury inherent in the pre-airbag design is significantly mitigated by introduction of the airbag, thus a reduction of the severity factor would be justified. This doesn't mean that the risk in a car without airbags has changed, however.