Carryover Effects in Repairable Systems

• Activity Number: 425
• Type: Contributed
• Date/Time: Tuesday, August 6, 2013 : 2:00 PM to 3:50 PM
• Sponsor: Section on Physical and Engineering Sciences
• Author(s): Candemir Cigsar*+
• Keywords: Imperfect repair ; internal covariates ; Poisson processes ; renewal Processes ; score tests

Abstract

The statistical modeling and analysis of data from repairable systems has been a major area of research in reliability and lifetime data analysis. Methods for analyzing adverse events (which we will term failures, for convenience) in systems have facilitated the analysis of patterns of failure, related factors, and the identification of opportunities for reliability improvement. Repairable systems reliability data typically involve recurrent events. An important aspect of the recurrent event processes is the presence of what we term carryover effects. By this we refer to settings where the intensity of an event is temporarily increased following the occurrence of an event. This phenomenon can occur when repairs to address a failure in a hardware or software system may not resolve the problem, or may even introduce new problems. Our objectives are to discuss models for carryover effects, and how evidence for or against their presence can be assessed. We provide simple tests for carryover effect in single or multiple processes. Asymptotic properties and the results of simulation studies concerning the tests are given. Methods are applied to a data set from industry.