What inspired this wish list request?
We often have reliability data with a mix of failure modes but during a point in development where those failure modes are not yet identified. But we still have a need to get an estimate of mean life of the overall distribution.
What is the improvement you would like to see?
When analyzing reliability data using the “Fit Competing Risk Mixture” model of the Life Distribution platform, it would be very helpful to have a way to determine the mean life (with confidence interval) of the resulting aggregated distribution.
Why is this idea important?
System level reliability data often has multiple failure modes but those failure modes are often unknown during development. These failure distributions aren't fit well by one parametric distribution, and the Fit Competing Risk Mixture is extremely useful. And while the profilers are powerful, often times we simply need to know the mean life of the aggregated distribution.
I think this may be very similar to the Mean Remaining Life Calculator in the Competing Cause report. We just don't know the causes yet.