Thursday, June 3, 2010

Reliability and safety analysis and systems are cheap- Failure is expensive

In the aftermath of the gulf oil spill I cannot sit by any longer. I work at NASA and I have been waiting for reliability data for 10 years for components that are used everyday. Even people who should know better argue about our reliability numbers, even after I say three times the numbers are assumed and look only at the relative percentage change- we end up talking about the basis for the numbers. Even within the PRA community people that I know that know statistics waste their time discussing numbers (and only mean numbers are mentioned- how about the uncertainty bounds?) I am guilty myself. Please stop it and stop it now. Probabilistic Risk Assessment is so powerful and it is not being used properly.

No comments: