A Bayesian Hierarchical Modelling Framework for Evaluating Transport Depot Maintenance System Adoption in Nigeria

{ "background": "The adoption of modern maintenance systems in transport depots is critical for infrastructure integrity and operational safety. Current evaluation methods often rely on aggregated metrics, failing to account for the hierarchical nature of data and regional heterogeneity, which limits the precision of adoption rate estimates and the identification of influential factors.", "purpose and objectives": "This article presents a novel Bayesian hierarchical modelling framework designed to rigorously evaluate the adoption rates of depot maintenance systems. Its objectives are to provide a robust method for quantifying adoption, to identify key technical and organisational drivers, and to produce region-specific estimates that inform targeted interventions.", "methodology": "A methodology is developed using a three-level hierarchical logistic model. The core model is specified as $\\text{logit}(p{ij}) = \\beta0 + \\beta X{ij} + uj + vk$, where $p{ij}$ is the probability of full system adoption for depot $i$ in region $j$, $X{ij}$ are depot-level covariates, $uj \\sim N(0, \\sigmau^2)$ are regional random effects, and $vk$ are sector-specific offsets. Inference is performed via Hamiltonian Monte Carlo sampling, with posterior distributions summarising all parameter and prediction uncertainty.", "findings": "The application of the framework to a case study demonstrates its utility, revealing that depot size and technician certification levels are the strongest positive predictors of adoption. A key quantitative finding is that the posterior probability of a positive effect for certified technician proportion exceeded 0.98, with a 90% credible interval for its odds ratio of [1.45, 2.10]. The model successfully quantified substantial regional variation, which was obscured in prior non-hierarchical analyses.", "conclusion": "The proposed Bayesian hierarchical model offers a statistically rigorous and practically superior methodology for evaluating maintenance system adoption. It formally incorporates data structure and uncertainty, yielding more reliable and actionable insights than conventional approaches.", "recommendations": "It is recommended that engineering assessments of system adoption