Ch. 30. Stochastic process models for reliability in dynamic environments

Nozer D. Singpurwalla, Thomas A. Mazzuchi, Süleyman Özekici, Refik Soyer

Research output: Chapters, Conference Papers, Creative and Literary WorksRGC 12 - Chapter in an edited book (Author)peer-review

3 Citations (Scopus)

Abstract

It is by now well recognized that the reliability of an item, be it hardware, software, or a biological unit, depends on the circumstances in which it is employed. Thus for example, a bug infested software that is never going to be used should be judged as being perfectly reliable! Hardware engineers refer to the circumstances of use as the "environment," whereas the software engineers use a more striking term, the "operational profile." There is much merit to this latter terminology. The word "profile" conveys the message that the circumstances of use change over time, and this in turn paves the way for stochastic process models to play a fundamental role in modern reliability theory. The aim of this chapter is to propose a framework for making the operational profile an integral part of reliability assessment. The framework is to be broad enough to include survival analysis as well. With the above in mind, we intend to develop a series of articles, each focusing on one aspect of the proposed framework. This chapter is on the stochastic foundations for describing and for incorporating the operational profile. Specialists in stochastic process theory will find little that is new here; it is intended to be an overview for those who may benefit from one, and as a vehicle to introduce some notation and terminology. © 2003 Elsevier Science B.V. All rights reserved.
Original languageEnglish
Title of host publicationStatistics in Industry
Pages1109-1129
Volume22
DOIs
Publication statusPublished - 2003

Publication series

NameHandbook of Statistics
Volume22
ISSN (Print)0169-7161

Fingerprint

Dive into the research topics of 'Ch. 30. Stochastic process models for reliability in dynamic environments'. Together they form a unique fingerprint.

Cite this