Real-Time Accident Detection and Data Analysis in Vehicles Using Data-Level Sensor Fusion, Machine Learning with LoRa, and Li-Fi Communication

Abstract

Motor vehicle accidents continue to be among the primary causes of mortality and serious injury across the globe. Traditional road accident detection and safety measures utilize either few sensors or a single mode of communication that might have a negative impact on the accuracy of identifying accidents and responding to emergencies. An Accident Detection and Data level sensor fusion is presented based on data-level sensor fusion and ML-based communication technologies. The recommended design incorporates (i) a range of sensors like the accelerometer, gyroscope, ultrasonic sensor, speed sensor, brake switch, and humidity sensor that constantly track vehicle dynamics and environment, (ii) machine learning systems that help process data from these sensors and detect collision incidents, (iii) Li-Fi communication technology for transferring alerts to other cars, and (iv) Long-range LoRa communications system used for notifying accident notifications and providing emergency alert information over long distances. Once there is any accident, then the system notes down the geolocation of the accident and provides information about that to rescue services and concerned people to help provide timely support. Data gathered is also used to analyze any accident to find out the reason behind the accident. The developed system uses techniques like data level sensor fusion, intelligent decisions, inter-vehicle communications, alerting of emergencies, and location tracking into one integrated framework.