AI-Based Smart Car Door Safety Mechanism Using YOLOv8 and Real-Time Object Tracking

Abstract

    Car door accidents, often referred to as "dooring," occur when a vehicle door is opened into the path of an oncoming two-wheeler or pedestrian, posing a significant threat to urban road safety. This paper presents an AI-powered smart car door safety system designed to restrict unsafe door opening through real-time object detection and automated mechanical intervention. The proposed architecture utilizes a YOLOv8 computer vision model integrated with an Arduino-based control unit and a central locking actuator. A side-mounted camera monitors a predefined 12-meter Region of Interest (ROI) to identify approaching risks. Upon detecting a potential collision within a critical 1.5-meter safety zone, the system automatically engages the vehicle's electronic locks and provides audio-visual warnings to the passengers. Experimental results demonstrate that the system achieves a detection accuracy of 90–92% with a response time of 0.8 to 1.2 seconds. This research provides a cost-effective, retrofittable solution to enhance passenger and pedestrian safety by bridging the gap between computer vision and active mechanical actuation.