When writing this book, it was suggested that I take a stab at de-mystifying Weibull Analysis. One valid comment I heard was," Although a lot of people use it, probably few people really understand it". Based on my experience, that statement is probably true.

Speaking for myself, I took only one course in statistics during my college career. While I received an A in the course, it could as easily have been a C or a D. All I really hoped to get out of the course was me. When it comes to statistical analysis, my mind reacts as it does to classical music. There is an on-off switch that immediately turns off, and it is a struggle to get it turned back on.

Contrary to my natural tendencies, I firmly believe that statistical analysis, and particularly Weibull analysis has an important role to play in reliability analysis and management. Weibull analysis is capable of predicting the rate and timing of future failures. More importantly, Weibull analysis enjoys the reputation of being accurate and reliable, so it is widely accepted. The fact that it is widely accepted is no small issue.

Assume, for example, that you have purchased 100 of the same kind of device (say an XB-22) or that you have purchased a large number of equipment items that contain the same device (an XB-22). Further assume that the failure of an XB-22 results in the loss of the income-producing capability so failures are unacceptable. Now assume...