Maintenance systems for transport depots are critical for infrastructure integrity and operational safety. In the Senegalese context, systematic evaluations of the methodologies underpinning these systems, particularly regarding long-term risk reduction, are lacking. This study conducts a comparative methodological evaluation of maintenance systems in the country's transport depots. Its primary objective is to quantify the efficacy of different methodological approaches in reducing structural and operational risks over an extended period. A comparative study employing multilevel regression analysis. The model, y₈₉ = ₀₉ + ₁₉x₈₉ + ₈₉, with ₀₉ = ₀₀ + ₀₁zⱼ + u₀₉, was fitted to longitudinal depot-level data, accounting for nested structures. Inference was based on robust standard errors to ensure reliability. The analysis reveals that depots implementing a predictive, data-driven maintenance methodology demonstrated a statistically significant 34% greater reduction in high-risk incident reports compared to those using reactive systems. The 95% confidence interval for this risk reduction ranged from 28% to 40%. Methodological choice in maintenance systems is a decisive factor for risk mitigation. The predictive approach offers a substantively superior pathway for enhancing long-term structural safety and operational reliability in transport infrastructure. Depot operators and regulatory bodies should prioritise investment in condition-monitoring technologies and staff training to enable a shift from reactive to predictive maintenance methodologies. National maintenance guidelines should be updated to reflect this evidence. infrastructure maintenance, risk analysis, multilevel modelling, predictive maintenance, transport engineering This paper provides the first application of multilevel regression modelling to comparatively evaluate maintenance methodologies in West African transport depots, introducing a novel framework for quantifying long-term systemic risk reduction.
Ndiaye et al. (Sun,) studied this question.