Low-Cost Predictive Modelling of Asphalt Mix Design Using Excel-Based Regression: A Sustainability-Oriented Engineering Approach from Nigeria

The need and quest for sustainable road infrastructure demands alternative, cost-effective and readily available tools that reduce environmental impacts during asphalt mix design. This study examines the application of Excel-based regression modelling as a cost-effective Artificial Intelligence tool to predict the optimum bitumen content in asphaltic concrete mixtures. Traditional mix design methods are resource-intensive and carbon-heavy, particularly in low-income countries. This study utilizes a dataset collected from six geopolitical zones in Nigeria and applies Multiple Linear Regression via Microsoft Excel to develop a predictive model. The model was calibrated and validated using standard error indices and physical lab tests. Results showed that the Excel-based model reduced the need for full-scale experimental mixes by over 90%, achieving an R² of 0.996 and a standard error of less than 0.35. This method significantly reduces material waste, emissions, and energy consumption. The study positions Excel-based regression modelling as a practical AI-enabled tool for engineers in resource-constrained environments. Future research should explore the integration of Excel with other add-in tools and the incorporation of real-time climate and traffic variables.