Driver fatigue and urban parking inefficiency are major contributors to road accidents and traffic congestion. This work presents Drive Guard, an integrated driver assistance system that combines computer vision and physiological monitoring for real-time detection of driver drowsiness and stress. The system utilizes dashboard-mounted cameras to analyze eye closure, facial expressions, and behavioral patterns, along with smartwatch-based heart rate (HR) and heart rate variability (HRV) data for physiological assessment. An ensemble learning approach combining Convolutional Neural Networks (CNN), Support Vector Machines (SVM), and k-Nearest Neighbors (k-NN) is employed for accurate classification of driver states. Additionally, an IoT-based intelligent parking module provides real-time parking availability and navigation support, reducing congestion and improving urban mobility. Experimental analysis demonstrates improved detection reliability and reduced false positives under real-world conditions. The proposed system offers a scalable and practical solution for enhancing driver safety and smart city infrastructure. This work is submitted as a preprint and is currently under consideration for further academic publication.
N et al. (2026) studied this question.