{ "background": "Transport maintenance depots are critical infrastructure for road network sustainability, yet systematic evaluations of their operational systems in low-resource settings are scarce. Existing studies often lack rigorous field-based methodologies to assess cost-effectiveness, leading to suboptimal asset management and budgetary allocation. ", "purpose and objectives": "This Data Descriptor presents a comprehensive dataset generated from a randomised field trial designed to methodologically evaluate and quantify the cost-effectiveness of different maintenance depot systems. The primary objective was to establish a robust empirical framework for comparing systematic versus ad-hoc maintenance regimes. ", "methodology": "A multi-arm, parallel-group randomised controlled trial was implemented across a network of depots. Depots were randomly assigned to either a centralised, technology-aided system or a conventional decentralised system. Data were collected on labour hours, spare parts consumption, vehicle downtime, and total operational expenditure over a standardised period. Cost-effectiveness was modelled using a generalised linear model: CEi = \0 + \1 Ti + \2 Xi + \, where CEi is a composite cost-effectiveness score for depot i, Ti is the treatment assignment, and Xi is a vector of covariates. Robust standard errors were clustered at the regional level. ", "findings": "The dataset reveals a clear directional trend favouring the centralised system. Analysis indicates a mean reduction in total operational expenditure of approximately 18% (95% confidence interval: 12% to 24%) for depots under the intervention, after controlling for fleet size and initial condition. The primary theme from ancillary qualitative data points to improved inventory management as a key driver. ", "conclusion": "The randomised trial methodology proved feasible for engineering management research in this context, generating a high-fidelity dataset for comparative analysis. The data structure enables robust inference on the economic and operational parameters of depot systems. ", "recommendations": "Future research should utilise this dataset to perform disaggregated lifecycle cost analysis. Practitioners
Uwimana et al. (Tue,) studied this question.