Review and classification of reliability measures for multistate and continuum models

Abstract In this paper we examine a variety of customer-centered reliability measures, generalized to multistate and continuum contexts. A classification scheme is proposed for these measures, and a new measure based on customer-usage patterns is defined. In each case, reliability measures are defined such that they are equally valid for binary, multistate, continuum, and mixed systems. Techniques which facilitate the use of continuum models are discussed, and time-dynamic, infinite-time, and partial-information modeling considerations are also addressed. Additional informationNotes on contributorsRUSSELL D. BRUNELLE Dr. Russell Brunelle received his Ph.D. in Industrial Engineering from the University of Washington (Seattle) in 1998. He received a B.S. in Physics and Mathematics from University of Puget Sound in 1992 and an M.S. in Industrial Engineering from University of Iowa in 1993. Russell Brunelle's research has included computer simulation of JIT manufacturing environments, Kalman-filtering analysis of arctic sea-ice evolution, stochastic modeling of student flow through undergraduate engineering programs, and multistate reliability analysis of aviation communication systems. His doctoral dissertation work was in the area of customer-centered multistate reliability models