Title: A geometric Brownian motion model for field degradation data
Author: E.A. Elsayed, H.T. Liao
Address: Department of Industrial and Systems Engineering, Rutgers University, 96 Frelinghuysen Road, Piscataway, NJ 08854-8018, USA. ' Department of Industrial and Systems Engineering, Rutgers University, 96 Frelinghuysen Road, Piscataway, NJ 08854-8018, USA
Journal: Int. J. of Materials and Product Technology, 2004 Vol.20, No.1/2/3, pp.51 - 72
Abstract: It is more difficult to estimate reliability when accelerated life testing (ALT) conditions are far from normal. Moreover, many of the test units do not fail when subjected to stresses close to their normal operating conditions during the test time. When components or products exhibit degradation behaviour, degradation testing may be an alternative to ALT to estimate reliability without subjecting them to actual failures. We propose a Geometric Brownian Motion Degradation Rate Model (GBMDR), which utilises field degradation data to estimate field reliability when the field stress context is unknown. A novel procedure for parameters estimation of the model is proposed. Its application is illustrated by utilising degradation data of a civil engineering structure to estimate field reliability.
Keywords: degradation model; field data; Brownian motion; reliability estimation.