APPLICATION OF ARTIFICIAL INTELLIGENCE IN FORECASTING DEMAND FOR CONSTRUCTION MATERIALS IN RWANDA

This research explores the application of Artificial Intelligence (AI) in forecasting demand for construction materials in Rwanda, aiming to enhance accuracy, efficiency, and scalability compared to traditional methods. A mixed-method approach was employed, combining qualitative insights from industry professionals with quantitative analysis of primary and secondary datasets. Neural networks emerged as the most effective model, achieving an R-squared value of 0.92, Mean Absolute Error (MAE) of 3,500, and Root Mean Square Error (RMSE) of 4,500. AI-based forecasts closely matched actual demand, maintaining errors within ±1.5%. However, adoption is hindered by limited expertise (r = -0.65, p < 0.05) and infrastructure constraints (r = -0.82, p < 0.05). The study concludes that AI can significantly optimize material forecasting, reducing inefficiencies and aligning Rwanda’s construction sector with global standards. Recommendations include capacity building, infrastructure investment, fostering public-private partnerships, and developing supportive policies to accelerate AI adoption.