Analysis of repairable system failure data using time series models

M. Xie, Siu Lau Ho

Research output: Journal article publicationJournal articleAcademic researchpeer-review

16 Citations (Scopus)

Abstract

Repairable system reliability analysis is very important to industry and, for complex systems, replacing a failed component is the most commonly used corrective maintenance action as it is an inexpensive way to restore the system to its functional state. However, failure data analysis for repairable systems is not an easy task and usually a number of assumptions which are difficult to validate have to be made. Despite the fact that time series models have the advantage of few such assumptions and they have been successfully applied in areas such as chemical processes, manufacturing and economics forecasting, its use in the field of reliability prediction has not been so widespread. In this paper, we examine the usefulness of this powerful technique in predicting system failures. Time series models are statistically and theoretically sound in their foundation and no postulation of models is required when analyzing failure data. Illustrative examples using actual data are presented. Comparison with the traditional Duane model, which is commonly used for repairable systems, is also discussed. The time series method gives satisfactory results in terms of its predictive performance and hence can be a viable alternative to the Duane model.
Original languageEnglish
Pages (from-to)50-61
Number of pages12
JournalJournal of Quality in Maintenance Engineering
Volume5
Issue number1
DOIs
Publication statusPublished - 1 Jan 1999
Externally publishedYes

ASJC Scopus subject areas

  • Safety, Risk, Reliability and Quality
  • Strategy and Management
  • Industrial and Manufacturing Engineering

Cite this