{ "background": "Transport maintenance depots are critical infrastructure for national road networks, yet their long-term cost-effectiveness is often poorly understood due to volatile operational expenditures and complex asset degradation patterns. In the South African context, a robust, data-driven forecasting framework is required for strategic budget allocation and lifecycle planning. ", "purpose and objectives": "This case study aims to methodologically evaluate the application of a time-series forecasting model to measure and predict the cost-effectiveness of transport maintenance depot systems. The objective is to assess the model's utility for long-term capital and operational expenditure planning. ", "methodology": "A case study methodology was employed, analysing historical operational and financial data from a representative national depot network. The core analytical tool was a seasonal autoregressive integrated moving average with exogenous variables (SARIMAX) model, specified as \ (B) \ (Bˢ) \ᵈ\D yt = \ (B) \ (Bˢ) \ + \ Xt, where Xₜ represents exogenous maintenance activity variables. Model diagnostics included analysis of robust standard errors and Ljung-Box Q-tests for residual autocorrelation. ", "findings": "The model demonstrated strong predictive capability for aggregate depot system costs, with a key finding being a forecasted gradual increase in the maintenance cost-to-asset value ratio of approximately 1. 2% per annum over the forecast horizon. Parameter estimates for labour and material cost drivers were statistically significant at the 95% confidence level, though forecast uncertainty bands widened notably beyond a five-year period. ", "conclusion": "The SARIMAX framework provides a technically sound and operationally useful methodology for forecasting depot system cost-effectiveness, enabling more informed infrastructure investment decisions. Its integration of exogenous operational factors offers a practical advantage over univariate models. ", "recommendations": "Infrastructure agencies should adopt similar time-series forecasting techniques within their asset management systems, ensuring consistent data collection on both financial and physical asset performance metrics to improve model calibration and long-term budget forecasts. ", "key words
Sipho Nkosi (Sun,) studied this question.