Efficient traffic management is essential for modern urban environments, where increasing vehicular density and dynamic traffic conditions pose significant challenges. This project aims to develop a real-time traffic flow prediction system leveraging advanced machine learning techniques to optimize traffic management, reduce congestion, and improve urban mobility. By integrating time- series forecasting, Long Short-Term Memory (LSTM) networks, and Graph Neural Networks (GNNs), the system will analyse historical and real-time traffic data, incorporating critical factors such as weather conditions, road infrastructure, and event occurrences. Key components include robust data pre-processing, feature engineering, and model training to capture complex spatial and temporal dependencies in traffic patterns. The system will address challenges such as handling missing data, ensuring scalability for large datasets, and adapting to unforeseen scenarios. Performance will be evaluated based on accuracy, latency, and robustness, enabling proactive traffic management strategies. This predictive capability will empower urban planners and traffic authorities to enhance road safety, reduce congestion, and improve urban mobility planning, contributing to more efficient and sustainable transportation systems.
Kumar et al. (Mon,) studied this question.