A Time-Series Forecasting Model for Evaluating Maintenance Depot System Adoption in Ethiopia (2000–2026)

The strategic adoption of centralised maintenance depot systems is critical for transport infrastructure resilience in developing economies. However, robust, quantitative methodologies for evaluating the long-term adoption and impact of such engineering systems are lacking, leading to suboptimal investment and policy decisions. This case study aims to develop and validate a time-series forecasting model to measure the adoption rate of centralised maintenance depot systems, providing a methodological framework for evidence-based engineering policy evaluation. A case study methodology was employed, utilising historical data on depot establishment, fleet size, and maintenance expenditure. A seasonal autoregressive integrated moving average (SARIMA) model, specified as SARIMA(1,1,1)(1,1,1)₁₂, was fitted to forecast adoption rates. Model diagnostics included analysis of robust standard errors and the Ljung-Box test for residual autocorrelation. The model forecasts a significant acceleration in system adoption, with the projected proportion of the national heavy vehicle fleet serviced by centralised depots increasing from an estimated baseline to over 60% within the forecast horizon. Parameter estimates for the autoregressive component were statistically significant (p < 0.01, robust SE = 0.15). The developed forecasting model provides a validated, quantitative tool for assessing the rollout of large-scale engineering maintenance systems, demonstrating that current policy frameworks are likely to achieve critical mass adoption. Infrastructure planners should integrate this forecasting methodology into long-term strategic asset management plans. Further research should focus on calibrating the model with real-time operational data from depot sensors. infrastructure management, maintenance engineering, time-series analysis, forecasting, adoption model, transport systems This study presents a novel application of SARIMA modelling to forecast the adoption trajectory of a national engineering system, providing a replicable method for evaluating infrastructure investment programmes.