Overview of Reliability Engineering

Abstract

The evolution of reliability since the late 1940 till 2000s is described. This evolution mainly came from United States, and the evolution is from system level reliability and maintenance to components reliability through statistical methods. Exponential distribution was used initially, and it was found not realistic in the 60s. Physics of failure began in the 90s, and reliability evaluation was then shifted from statistical methods to physics of failure. Since 2000, the hybrid physics–statistical approach evolve due to the need to combine both of them in order to evaluate component and system reliability realistically. While reliability is evolving and maturing in USA, the concept of reliability was introduced to Asia only in the late 70s in the form of qualification as US manufacturers were shifting their production to Asia, and they needed to qualify the factories in Asia. Unfortunately, the growth of reliability methodology is very slow in Asia, and still many are using exponential distribution to evaluate reliability. On the other hand, physics of failure approach is growing fast in Asia, due probably to the need to troubleshoot failures manufactured in Asia. This chapter helps the readers to have a border view of reliability and the necessity for its evolution.