{ "background": "Time-series forecasting is a critical tool for infrastructure asset management, yet the methodological robustness of models applied in West African contexts is seldom scrutinised. A previously published autoregressive integrated moving average (ARIMA) model for forecasting transport depot system reliability in a West African nation required independent validation.", "purpose and objectives": "This study aimed to conduct a direct replication and methodological evaluation of the specified ARIMA forecasting model. The objectives were to assess the reproducibility of the model's parameter estimation and forecast accuracy, and to evaluate its methodological rigour against contemporary standards.", "methodology": "We executed a direct computational replication using the originally specified methodology and subsequently expanded the model form to include exogenous maintenance expenditure, testing $yt = \\beta0 + \\beta1 y{t-1} + \\beta2 xt + \\epsilont$ where $xt$ is the exogenous variable. Model diagnostics, including residual analysis and out-of-sample forecast errors, were compared using robust standard errors.", "findings": "The replication confirmed the original model's core seasonal pattern but revealed a systematic over-forecast of reliability by approximately 8% in medium-term projections. Incorporating exogenous maintenance data significantly improved model fit, reducing the mean absolute percentage error by 15 percentage points in validation.", "conclusion": "The original model's structural form is reproducible but exhibits a positive bias in forecasting. The methodological evaluation underscores the sensitivity of reliability forecasts to maintenance inputs, a factor omitted from the initial specification.", "recommendations": "Future forecasting efforts for depot reliability should explicitly incorporate maintenance expenditure as a covariate. Practitioners should apply the expanded model form and conduct regular out-of-sample validation to mitigate forecast drift.", "key words": "replication study; time-series forecasting; system reliability; transport infrastructure; maintenance engineering; ARIMA", "contribution statement": "This study provides the first independent validation and methodological extension of a key reliability forecasting model for West African transport depots, demonstrating the critical importance of including maintenance covariates