This paper introduces a predictive traffic accident warning system tailored for smart city applications, aiming to improve urban safety through advanced data analysis and contextual reporting. The proposed system employs computationally efficient algorithms to predict the severity of traffic accidents with high accuracy while maintaining robust data privacy standards. By integrating real-time traffic data with external knowledge sources, it generates detailed, actionable reports and timely warnings, enabling proactive decision-making. The system’s design emphasizes effective task orchestration for seamless integration with existing urban infrastructure and optimized resource management. Evaluation results demonstrate the system’s high accuracy, scalability, and practical viability in smart city environments. Future research will focus on enhancing model efficiency, leveraging transfer learning for cross-domain adaptability, and deploying real-world implementations to vali- date the system’s performance.
Sai Batchu (Tue,) studied this question.