Reliability Engineering for High-Value Low-Volume Complex Equipment

Semiconductor manufacturing equipment is one type of high-value low-volume products. Each piece of such equipment could cost millions of dollars, but volume could be in low hundreds yearly. Such equipment involves mechanical engineering, electrical engineering, software and controls, chemistry delivery, RF power delivery, etc., which make the system very complicated and vulnerable to reliability issues. On the other hand, due to Moore's law, such equipment gets upgraded very fast, and customers constantly push for new product delivery. The reliability data acquisition becomes very difficult in terms of both time and cost. However, there are very few publications in the literature about how to systematically apply reliability engineering to high-value low-volume products. In this paper, we will focus on how to conduct reliability engineering work based on the authors' years of experience in semiconductor manufacturing equipment industry. Due to the characteristics of semiconductor manufacturing equipment, it is concluded that reliability engineering should focus on design for reliability (i.e., reliability qualification plan, system reliability assessment, failure mode and effects analysis) at the product development phase, and utilize knowledge and expertise sharing from existing products and coordination with the vendors. In addition, reliability demonstration at subsystem and system level will be discussed. Finally, close loop reliability feedback and Continuous Improvement Plan (CIP) are curial for the reliability growth, which can be demonstrated by Duane plot.