A Locally Autonomous Driving System Driven by Collaborative AI Vision and IoT Monitoring for Path Planning and Decision-Making

Abstract

This paper presents a novel domestic autonomous driving system design that synergistically integrates AI vision and IoT monitoring technologies. The proposed platform adopts a multi-sensor fusion architecture, combining edge computing with lightweight model deployment to establish a robust closed-loop "perception-decision-control" system. At the core of the system lies the YOLOv5 object detection model, which enables real-time and accurate identification of dynamic obstacles. Enhanced by Beidou-3/GPS dual-mode high-precision positioning, the system ensures precise localization even in challenging environments. Furthermore, 5G/Beidou short-message dual-channel communication facilitates seamless data transmission, enabling remote monitoring and control. Through these advanced technologies, the system achieves dynamic obstacle avoidance, adaptive speed regulation, and efficient route planning, significantly enhancing operational safety and efficiency. The solution offers a highly reliable and cost-effective domestic alternative for industrial autonomous driving. By leveraging AI vision and IoT monitoring, the system reduces reliance on expensive proprietary sensors and infrastructure, making it accessible to a broader range of applications. The closed-loop design ensures rapid response to changing environmental conditions, while the lightweight model deployment minimizes computational overhead. This innovative approach not only addresses the challenges of autonomous driving in complex industrial settings but also provides a scalable framework for future advancements. The proposed system represents a significant step forward in the development of affordable and efficient domestic autonomous driving solutions.