ATP 5-19 "Risk Management" 2019-11-17

[Bill Levinson]

Bill Levinson submitted a new resource:

ATP 5-19 "Risk Management" - U.S. Army's Risk Management Process from 2014

This is in the public domain as a publication of the U.S. Government, and bears many similarities to ISO 30001.

It brings up a point not raised in FMEA; that the risk of occurrence (ATP 5-19 calls it frequency) isn't just the individual chance of occurrence, it also reflects the frequency with which we are exposed to the risk, e.g. how many times we do the job, or how many parts we make.

It can be used with DD2977 (Deliberate Risk Assessment) which is simpler than traditional FMEA (if...

Read more about this resource...

 

[toniriazor]

Hello thank you for interesting information.
Sounds interesting the part with the frequence - I never thought about this just what is the likelihood a failure to occur or in other words only about what FMEA is offereing. I guess frequence can be applied in regards to occurence and detection ?
What about frequency calculation ? how do you work it out ?

 

[Bill Levinson]

Hello thank you for interesting information.
Sounds interesting the part with the frequence - I never thought about this just what is the likelihood a failure to occur or in other words only about what FMEA is offereing. I guess frequence can be applied in regards to occurence and detection ?
What about frequency calculation ? how do you work it out ?

DD2977 is more subjective than the old FMEA (which matched probabilities of occurrence with the FMEA ratings). The principle was that risk is proportional not just to the individual chance of occurrence, but also the frequency with which we are exposed to the risk. The takeaway is that, unless the chance is zero, the undesirable event will occur if we perform the job enough times, which means we have to reduce the chance to zero with error-proofing or engineering controls instead of relying on administrative controls that require compliance and vigilance.

The new AIAG/VDA FMEA manual, though, assigns occurrence ratings based not on a quantified chance of occurrence, but rather based on whatever preventive controls are in place. The only way to get a 1 occurrence rating is to make the undesirable event impossible.