"background": "Municipal infrastructure asset management in South Africa requires robust predictive tools to plan for maintenance and capital expenditure. A previously proposed time-series forecasting model for infrastructure efficiency gains has been cited in policy discussions but lacks independent validation of its methodological rigour and predictive performance. ", "purpose and objectives": "This study aimed to replicate and critically evaluate the methodological framework of the specified forecasting model. The primary objective was to assess its computational reproducibility, statistical validity, and sensitivity to input assumptions. ", "methodology": "The original model specification, yt = \0 + \1 (A{t-1) + \2 Tt + \, where A represents asset condition and T is a time trend, was re-implemented using open-source software. The replication process tested the model's stability by applying bootstrapping techniques to derive robust standard errors and conducting sensitivity analyses on key parameters. ", "findings": "The replication confirmed the model's core computational output but revealed a significant sensitivity to the initial asset condition parameter; a 10% variation in this input altered long-term efficiency forecasts by approximately 15%. The bootstrapped confidence intervals for the trend coefficient \2 were substantially wider than originally reported, indicating greater forecasting uncertainty. ", "conclusion": "While the model is computationally reproducible, its utility for long-term strategic planning is limited by its sensitivity to initial conditions and the under-reported uncertainty in its trend estimates. ", "recommendations": "Future applications of this model should incorporate probabilistic forecasting methods and more comprehensive uncertainty quantification. Municipal engineers should treat its point forecasts with caution, using them as one input within a broader scenario-based planning framework. ", "key words": "asset management, forecasting, infrastructure, replication, time-series, uncertainty quantification", "contribution statement": "This study provides the first independent, methodological validation of a cited forecasting model, offering a critical technical assessment of its limitations for infrastructure planning and establishing a revised protocol for
Naidoo et al. (Thu,) studied this question.