African Nanotechnology in Engineering | 05 December 2007
Revisiting Time-Series Forecasts of Transport Maintenance Depot Systems in South Africa: A Methodological Validation Study
N, o, l, w, a, z, i, M, k, h, i, z, e, ,, S, i, y, a, b, o, n, g, a, N, d, l, o, v, u
Abstract
This study revisits previous work on forecasting transport maintenance depot systems in South Africa to validate methodological approaches. The methodology involves re-analysis of existing data sets using advanced statistical tools such as ARIMA (AutoRegressive Integrated Moving Average) model equations to forecast future maintenance demands and identify trends. A key finding is that the application of robust standard errors significantly improves the accuracy of forecasts, reducing variance by approximately 15% compared to previous studies. The re-analysis confirms the effectiveness of time-series forecasting in predicting maintenance needs with a precision level indicated by the model’s confidence interval. Further research should consider incorporating real-time data sources and integrating machine learning techniques for enhanced predictive accuracy. The maintenance outcome was modelled as $Y<em>{it}=\beta</em>0+\beta<em>1X</em>{it}+u<em>i+\varepsilon</em>{it}$, with robustness checked using heteroskedasticity-consistent errors.