Safety in e-Scooters: a Machine-Learning Approach for Online Second Passenger Detection

Electric scooters are pivotal for urban mobility, yet safety concerns still prevent widespread adoption. Research identifes rider negligence, such as carrying a second passenger, as a major risk. To address this issue, we propose an autonomous system for real-time detection of second passengers. By analyzing vehicle dynamics through minimal sensors and employing an interpretable machine learning approach, our solution ensures accuracy and interpretability. Rigorous testing with diverse users validates its effectiveness, showcasing adaptability to user characteristics and road conditions, proving the potential of this approach to foster safer electric scooter usage.