{ "background": "The reliability of railway maintenance depot systems is critical for operational continuity and safety, yet quantitative assessment methods tailored to the specific operational and environmental conditions of developing networks are lacking. Existing reliability models often fail to account for hierarchical data structures and inherent uncertainties in maintenance processes. ", "purpose and objectives": "This article presents a novel Bayesian hierarchical modelling framework for the reliability assessment of railway maintenance depot systems. The primary objective is to provide a robust methodology that integrates multi-level operational data to quantify system reliability and identify key influencing factors. ", "methodology": "A three-level Bayesian hierarchical model is developed, where the failure rate ij for component i in depot j is modelled as ij \ (\, \), with depot-level parameters \, \ drawn from a common hyperprior distribution. Inference is performed using Hamiltonian Monte Carlo sampling, with posterior credible intervals quantifying parameter uncertainty. ", "findings": "The model application to case study data indicates that depot-level operational practices explain approximately 40% of the variance in component failure rates, a substantially greater proportion than component age alone. The 95% highest posterior density interval for the hyperparameter governing mean time between failures across all depots was 145, 189 hours. ", "conclusion": "The proposed Bayesian hierarchical model provides a statistically robust and operationally informative framework for reliability assessment, effectively handling data heterogeneity and uncertainty quantification in maintenance systems. ", "recommendations": "Adoption of this modelling approach is recommended for infrastructure managers seeking to move from reactive to predictive maintenance strategies. Future work should focus on integrating real-time sensor data into the model's hierarchical structure. ", "key words": "Bayesian inference, hierarchical modelling, reliability engineering, maintenance optimisation, railway infrastructure", "contribution statement": "This paper introduces a novel Bayesian hierarchical model specifically configured for the reliability analysis of maintenance depot systems, demonstrating its utility through a detailed case study on Ethiopian railway depots. "
Assefa et al. (Tue,) studied this question.