Enhancing Road Safety with Connected Internet of Things (IoT) based Vehicle Technology using Cooperative Hazard Awareness and Avoidance Scheme

The increasing demand for road safety in modern transportation systems has led to the integration of Internet of Things (IoT)-based vehicle technology. This paper proposes a Cooperative Hazard Awareness and Avoidance (CHAA) scheme that leverages IoT, reinforcement learning (RL), and gradient boosting machines (GBM) to enhance realtime hazard detection and avoidance. The system integrates multiple sensors, including LiDAR, RADAR, and cameras, to collect real-time data on road conditions, surrounding vehicles, and pedestrians. The proposed RL+GBM model is compared with nine existing machine learning models, such as CNN, LSTM, and XGBoost. The proposed model achieves an accuracy of $94.8 \%$, outperforming the next-best model, CNN, which achieved $91.2 \%$ accuracy. Precision, recall, and F1-score were also higher for the proposed model, with values of $92.4 \%$, $\mathbf{9 1 . 6 \%}$, and $\mathbf{9 2 . 0 \%}$, respectively. Additionally, the model demonstrates superior energy efficiency and lower latency, making it well-suited for IoT environments where real-time processing is critical. These results indicate that the CHAA system is a robust solution for enhancing road safety, offering significant improvements over existing methods in terms of accuracy, responsiveness, and energy consumption.