The IEC 61508 standard states that automatic diagnostics can only lower the dangerous undetected failure rate if those diagnostics execute 100 times faster than the expected demand rate. That requirement has always seemed excessive to many of us here at exida. As an accredited Certification Body we must enforce that requirement unless we can prove it to be excessive by deriving the basis for a new/modified requirement and present those results in a peer reviewed, journaled paper. Only then can we update the exida scheme to use better requirements.
Last year Dr. J. Bukowski began work on this problem and successfully derived the analytic formulas for correct diagnostic credit. It seems a strong partial credit (78%) can be given for only a 2X faster diagnostic rate. That matches the sampling theory taught in the university. Great news. The derivation and the results will be presented in a paper at the Reliability and Maintainability Symposium (RAMS) on January 23, 2017.
Anyone who wants to see preliminary results can also read the exida White Paper, “Properly Assessing Diagnostic Credit in Safety Instrumented Functions Operating in High Demand Mode” on the exida website.