I am sorry for having misunderstood what you said. I use to have access to a data base where moments of failure detection are registered over the years for each particular part and failure mode. My colleagues estimate remnant life based on the mechanics of failure (considering the state of the material and the specific working conditions) in order to advice the owner of the premise when to stop and replace or perform one more inspection to confirm or reschedulle. Whenever needed, they also estimate the so called "incubation period of the failure" in order to have a better picture of where the P point lies in a time scale.