Conference proceeding
CBM Policy Optimization Based on Lévy Copula Modeling of Latent Degradation
Proc. of the European Safety and Reliability Conference (ESREL2026), pp.2508-2514
The European Safety and Reliability Conference (ESREL 2026) (University of Minho in Braga, Portugal, 14/06/2026–19/06/2026)
2026
Abstract
Within practical systems, the continuous buildup of wear and tear is an unavoidable reality. Industries subject to such systems must account for degradation to maximize runtime while minimizing costs. Among common methods, Condition-Based Maintenance (CBM) uses observable system conditions to make decisions. This is considered both cost effective and reliable with regard to planning system maintenance. Latent degradation refers to any degradation that is hidden within a system and can cause breakdowns without warning. CBM can use correlated performance characteristics, called markers, to make predictions about latent degradation for decision making. Recent evidence suggests that a copula-based approach may be better at capturing the dependence structure between latent degradation and markers compared to previous methods. This study investigates the use of the Lévy copula function to model the dependence between observed markers and latent degradation, enabling prediction of latent degradation from system markers.
Details
- Title
- CBM Policy Optimization Based on Lévy Copula Modeling of Latent Degradation
- Authors/Creators
- Connor Stewart-Green - Murdoch University, School of Mathematics, Statistics, Chemistry and PhysicsSoudabeh Shemehsavar - Murdoch University, School of Mathematics, Statistics, Chemistry and PhysicsGraeme Hocking - Murdoch University, School of Mathematics, Statistics, Chemistry and Physics
- Publication Details
- Proc. of the European Safety and Reliability Conference (ESREL2026), pp.2508-2514
- Conference
- The European Safety and Reliability Conference (ESREL 2026) (University of Minho in Braga, Portugal, 14/06/2026–19/06/2026)
- Publisher
- Research Publishing
- Identifiers
- 991005904678007891
- Copyright
- © 2026 ESREL 2026 Organizers.
- Murdoch Affiliation
- School of Mathematics, Statistics, Chemistry and Physics
- Language
- English
- Resource Type
- Conference proceeding
Metrics
1 Record Views