Development of a Predictive Model for Maintenance Delays Impact on Gravel Road Life-Cycle Costs in Tanzania

Abstract

This study developed a predictive model for evaluating the impact of road maintenance delays on the life-cycle cost of gravel roads in Kinondoni District, under the jurisdiction of the Tanzania Rural and Urban Roads Agency (TARURA). Gravel roads play a vital role in supporting socio-economic activities in both rural and urban areas of Tanzania. However, delayed maintenance has continued to pose significant challenges, leading to road deterioration, higher long-term costs, and reduced accessibility. A mixed-methods approach was employed, combining qualitative data obtained through interviews with TARURA officials and contractors, and quantitative data collected from 44 road segments representing diverse geographic and traffic conditions. The study analysed maintenance records spanning five years (2019-2024) and incorporated economic data from national road fund allocations. The Relative Importance Index (RII) was used to rank delay factors, while a Multiple Regression model was manually developed in Excel to predict life-cycle costs based on delay severity. Statistical validation included correlation analysis, residual diagnostics, and cross-validation using independent datasets. Key findings revealed that out of twelve assessed delay factors, seven were significant predictors of cost increases. These include inadequate funding release (RII = 0.941), multiple damage points (RII = 0.891), poor traffic management (RII = 0.850), equipment availability (RII = 0.809), material transport issues (RII = 0.782), emergency repair prioritisation (RII = 0.768), and rainfall intensity (RII = 0.759). The life-cycle cost model yielded an R² value of 0.8735, indicating strong predictive capability with 87.35% of cost variance explained by delay factors. Model validation using independent road segments showed prediction accuracy within 15% of actual costs in 82% of cases. The study concludes that maintenance delays significantly affect the long-term sustainability and cost-efficiency of gravel roads, with each year of delay increasing life-cycle costs by approximately 0.62 million TZS. It recommends strengthening planning mechanisms, improving budget disbursement efficiency, enhancing coordination of maintenance activities, and adopting predictive modelling in decision-making. The models developed provide practical tools for TARURA and other road agencies to prioritise interventions, forecast costs, and optimise road asset management strategies.