"background": "The performance of transport maintenance depots is critical for infrastructure integrity and economic development. In Ethiopia, systemic inefficiencies persist, but a comprehensive, quantitative methodological framework for evaluating these systems and their adoption drivers is absent. ", "purpose and objectives": "This study aims to develop and apply a novel methodological framework for evaluating depot systems, with the primary objective of quantifying the factors influencing the adoption rates of structured maintenance protocols. ", "methodology": "A multilevel regression model was constructed, nesting depot-level observations within regional administrative units. The model, ij = \0j + \1X{ij + uj + eij, where \0j = \00 + \01Zj + v₉, was estimated using restricted maximum likelihood with robust standard errors. Data were collected via a stratified survey of depot managers and technical audits. ", "findings": "The multilevel analysis revealed that technician training levels and inventory digitalisation were the strongest positive predictors of protocol adoption. A one-standard-deviation increase in training investment was associated with a 17. 2 percentage point increase in adoption likelihood (95% CI: 12. 4 to 22. 0). Regional institutional support showed no statistically significant effect at the 5% level. ", "conclusion": "The methodological framework successfully isolates depot-level operational factors as the primary drivers of system adoption, overshadowing broader regional policy influences. This indicates that improvement initiatives should be targeted at the operational unit. ", "recommendations": "Depot management should prioritise investment in continuous technician training and the implementation of digital inventory systems. Policymakers are advised to reorient support mechanisms to directly enable these depot-level capabilities. ", "key words": "infrastructure management, maintenance systems, multilevel modelling, regression analysis, adoption drivers", "contribution statement": "This paper provides a novel hierarchical modelling framework for infrastructure system evaluation and presents the
Hailu et al. (Fri,) studied this question.