Behavior-Aware Real-Time Navigation for Safer Driving Routes

This paper presents a real-time framework for analyzing safe driving behavior and the interaction of moving vehicles in traffic scenes. Predicting the safety paths of both the driver’s vehicle and nearby vehicles is essential for enhancing road safety. Such predictions can assist the driver in better understanding the driving environment and anticipating the intentions of neighboring drivers. This work focuses on real-time prediction of driver behavior using machine learning techniques applied to data collected from smartphone sensors (accelerometer, gyroscope, GPS) and OBD II. To achieve real-time performance, we implemented a real-time architecture leveraging Atlas MongoDB to synchronize data communication. Additionally, our framework includes a safety recommendation system that helps drivers finding safer routes by avoiding aggressive or slow driving behaviors.