Customer Reviews

Statistical Methods for the Reliability of Repairable Systems

5 star:		(1)
4 star:		(1)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

Most reliability curves or their survival curve counterparts analyze time to failure without considering that there may be a difference in distribution for a part that was repaired compared to one that is a perfectly new replacement or they assume that the part being analyzed is not repairable. But we know with automobiles and other complex systems there are parts that are usually replaced and other parts that are usually repaired. Sometimes you have a choice and usually it is economic considerations that dictate the decision. In practice repairable systems are common, airplanes, cars, trains, and televisions to name a few. Of course very simple items are designed not to be repairable with the light bulb being an obvious example. So there should be much more written about them. In both the reliability and survival literature the theory for repairable systems is maturing. The theory of recurrent events that is presented in fine books such as the one by Wayne Nelson and the even more recent text by Cook and Lawless. These methods are sometimes applicable to repairable systems. In his book Nelson refers to Harry Ascher's text as the only one available on the topic of repairable systems.

This text is a welcome current addition to the statistical methods and models used to address practical problems about systems that have repairable parts.

This is a great reference book on repairable systems reliability. Too many reliability engineers, and especially their management, blindly rely on "Weibulls" for decision making. Often the subtle (and critical) difference between a non-repairable component models (like the usual Weibull & Lognormal models) and a repairable-system models are overlooked. This book clearly explains the differences between repairable & non-repairable systems, minimal-repair and renewal models, etc. It has a nice balance between theory and examples, and one of the best explanations of poisson process theory I have seen.

I have applied the methods described in this book to several industry data sets (from my own experience dealing with gas & steam turbine power plants, wind turbine farms, control systems, etc) and they work very well. And the theory finally helped clarify a lot of questions I had while using commercial reliability growth programs and standards.

The proofs can be a bit intimidatng for the novice, but its worth chugging through the math as some of the most useful (and practical) results are found in the lemmas. I would rate this as a classic for repairable systems reliability, just as Nelson's "applied life data analysis" is the classic for non-repairable systems. Both books should be on the bookshelf of any industrial or reliability engineer.