[PDF]

AI-Based Urban Traffic Congestion Prediction and Optimization Model


Brandon Williams

08/05/2025

Supervised by Yipeng Qin; Moderated by Liam Turner

Traffic congestion is a major issue in Cardiff, causing delays, pollution, and inefficiencies in transport planning. My project aims to develop an AI-based system to predict and reduce congestion by identifying traffic hotspots using OpenStreetMap data and machine learning. By analysing historical traffic data, the system will suggest better travel routes to ease congestion. This approach involves processing traffic data, applying predictive models, and testing optimization strategies to improve traffic flow. The project will help make Cardiff’s roads more efficient and contribute to smarter urban planning.


Initial Plan (03/02/2025) [Zip Archive]

Final Report (08/05/2025) [Zip Archive]

Publication Form