Accelerated Testing: Statistical Models, Test Plans, and Data Analyses

The author comes from GE background which did extensive accelerated testing on a variety of product types. Introductory material in advanced statistical modeling of acceleration mechanism using different distribution functions: exp, gauss, weibull, and others. Excellent blend of math, statistics, and practical experimental techniques. A great reference book.

This is undoubtedly the best book in quantitative accelerated life testing. Dr.Nelson does an excellent job in clearly explaining the statistical models and the life data analysis concepts related to accelerated testing. I cant think of any other book that comes even close to this work. A great reference for anyone interested in Accelerated Testing.

Solid reference for both beginners and experienced readers. Author provides great synonyms for readers from various backgrounds.

Compared to what you can found for free on the web today the content has to be improved or at least have a companion website to download test files or detailed mathematical proofs.

Nelson is a private consultant who has worked on countless practical problems in reliability from his consulting practice and previous employment at General Electric. He is an excellent lecturer and writer. His earlier book "Applied Life Data Analysis" was considered to be one of the best texts on reliability.This book is very thorough in its treatment of all aspects of accelerated testing and is filled with many good references. Nelson carefully defines the mathematical models which consist of two components, (1) an acceleration function which describes how the mean lifetime changes as a function of the acceleration factor and (2) a probability distribution that explains the random variability of outcomes at each acceleration factor. A particular mean function could be the Arrhenius relationship and the probability distribution could be exponential. Hence there is not a single Arrhenius acceleration model but rather an Arrhenius-exponential, an Arrhenius-lognormal or an Arrhenius-Weibull model. The book is filled with interesting theory and examples. Nelson provides excellent practical guidance based on his wealth of experience.

Dr. Taylor does his usual great job covering a difficult subject - ties in well with his "Applied Life Data Analysis", the revered reliability resource. Because of his GE background, he emphasizes experimental plans and analysis methods (i.e.: partially censored) which permit you to make good predictions regarding a product or system while you are still collecting data (e.g.: management wants an answer NOW, even if the experiment is not completed). A few more examples of how "typical" systems behave would be helpful, but otherwise a great resource.

I agree with the previous reviewer that this is the definitive classic in accelerated life testing. It's a great companion volume to Dr. Nelson's "Applied Life Data Analysis" which remains one of the best reliability textbooks ever written. I have noticed that all major reliability and accelerated testing programs use this book as their main reference - proof of the books usefulness. My only compaint is with the publishers for the steep price - now that they're bringing out a lower cost version of Applied Life Data Analysis, I hope they'll bring out a low-cost or paperback version of this classic as well.

I had been using my library's copy of this book for a while but just liked the book so much that I had to commit on buying it. Nelson provides a thorough and yet very applied manual which covers the basics and provides a lot of practical examples and tips. It gives a lot of different views and ways to come to your ends, which makes things all the more interesting. This book has proved a valuable companion in my research.However, even though the book allows for enough review of basic statistics to get along, it won't replace a good book on the topic and you may want to get one as a sidekick.