Predicting Failures of the Brazilian Air Force Tucano Fleet Using Survival Analysis
Abstract
During the life of an aircraft, several failures may occur causing unplanned maintenance and costs of parts, labor, transportation, lost opportunities, and operations, among others. The prediction of these events becomes especially difficult for new airframes, since they do not have many flight hours or enough data to observe failures. This study analyzes and model the time between failures (TBF) of the Brazilian Air Force T-27 Tucano fleet by applying various non-parametric, semi-parametric, and parametric statistical models, to the TBF, such as the descriptive statistics, Kaplan-Meier estimators, Cox proportional hazards models, with or without frailty, and survival regression models, with or without frailty. The study concludes by proposing a failure model that can be applied to the new similar airframes.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 01, 2019
- Accession Number
- AD1077518
Entities
People
- Luciana M. Monteiro
Organizations
- Air Force Institute of Technology