SRC Forum - Message Replies
Forum: Reliability & Maintainability Questions and Answers
Topic: Reliability & Maintainability Questions and Answers
Topic Posted by: Reliability & Maintainability Forum
Organization: System Reliability Center
Date Posted: Mon Aug 31 12:47:36 US/Eastern 1998
Posted by: Sadat
Date posted: Thu Jul 28 16:56:00 US/Eastern 2005
Subject: Early Life Prediction
I am wondering if anyone has done any early life failure rate predictions. If yes, then what methodology did you use?
Telcordia SR-332 has First Year Multiplier (FYM) operating on the steady-state failure rate. I need to predict the first year and second year failure rates based on my long-term MTBF number.
Subject: Credible failure rate prediction
Reply Posted by: Larry George
Organization: Problem Solving Tools
Date Posted: Sun Jul 31 13:22:34 US/Eastern 2005
I've made age-specific predictions of failure rates, including DoA and early life failure rates. Use observed, age-specific field failure rates of comparable products or parts scaled by the ratio of old and new MTBF predictions. This method was inspired by the observation that, although designs change, reliability doesn't change, much, and the change in reliability is approximated by the ratio of MTBF predictions, which largely capture parts count differences. The predictions were pretty well confirmed by later experience, although one product had more infant mortality than it's predecessors. We believed the prediction well enough to investigate the excessive infant mortality.
The method is a simple adaptation of proportional hazards or relative risk model, which has widespread use in biostatistics, http://rac.alionscience.com/pdf/4Q2003.pdf.
For more information, see http://www.asq.org/reliability/pub/index.html or my web site.