Design of an IoT-based Vehicle Accident Severity Detection System with Artificial Intelligence

Traffic accidents are a significant global issue with severe human and economic repercussions, demanding rapid and precise emergency response. This paper presents an Internet of Things (IoT)-based accident severity detection system aimed at enhancing the speed and accuracy of emergency response through real-time data analysis. The system integrates a set of sensors on a remote-controlled vehicle to monitor critical indicators of accidents, transmitting data via two communication channels: a GSM module for immediate SMS alerts to a designated contact and an ESP32 WiFi module for continuous data streaming to Node-RED. Utilizing machine learning, specifically a Random Forest model selected for its superior accuracy of 80.7%, the system classifies accident severity effectively. This IoT framework, by optimizing resource allocation and enabling timely intervention, offers a promising advancement in emergency response strategies, potentially reducing fatalities and injuries associated with traffic incidents. The proposed system exemplifies how IoT and machine learning can drive innovations in road safety, setting a foundation for intelligent accident detection and communication in future transportation systems.