Activity Number:
|
79
- Contributed Poster Presentations: Lifetime Data Science Section
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 3, 2020 : 10:00 AM to 2:00 PM
|
Sponsor:
|
Lifetime Data Science Section
|
Abstract #309901
|
|
Title:
|
Prediction of Future Failures for Heterogenous Reliability Field Data
|
Author(s):
|
Colin Lewis-Beck* and Qinglong Tian and William Meeker
|
Companies:
|
University of Iowa and Iowa State University and Iowa State University
|
Keywords:
|
Reliability ;
Bathtub hazard;
Censored data;
Bayesian;
Prediction
|
Abstract:
|
This paper introduces a Bayesian approach for using lifetime data to construct prediction intervals for future failures. We first present the generalized limited failure population (GLFP) model. This five-parameter model for lifetime data accommodates lifetime distributions with multiple failure modes: early failures (sometimes referred to in the literature as “infant mortality”) and failures due to wearout. We fit the GLFP model to a heterogenous population of lifetime data using a hierarchical modeling approach. We then predict the number of future failures for each sub-population using an approximation of the Poisson-binomial distribution. We evaluate our prediction intervals using a small simulation study, and apply our methods to real field data on hard drive reliability.
|
Authors who are presenting talks have a * after their name.