An Enhanced Alert System for Accidents Involving Electric Vehicles

Abstract

   This paper introduces the development of an integrated data-based Electric Vehicle Accident Alert System (EVAAS), which is powered by Machine Learning, the Internet of Things(IoT), and cloud computing for better road safety & emergency handling by electric vehicles. The ESP8266-based system processes real-time GPS data using a K-Nearest Neighbors (KNN) classification algorithm to determine the likelihood of accidents in particular regions. When a crash happens, the EVAAS system will automatically send an SOS alert with accurate positioning data to one's pasted emergency contacts. EVAAS features a modular, plug-and-play design with an intuitive web interface for device registration and cloud integrations. Every device has a unique identifier linked to it on registration, so setup is easy and if need be quickly deployed in cars. Information from accidents is recorded continuously, and accident locations that are found get included in the ML model to increase prediction accuracy over time (the system teaches itself). This involves the use of Google Firebase for cloud services to enable enterprise-grade scalable and consistent data management. By converging the powers of IoT, ML and cloud technologies, we deliver resilient technology that is adaptable to changing EV-specific safety requirements, offering a human-centred approach. This paper also presents the architecture of EVAAS; model developments, IoT, and cloud integration are the focus of this study. We will also show performance measurements, detailing how accurate/fast/and scalable the system is. We end by discussing potential future improvements and externalities of such systems to enhance road safety in the era of electric mobility.