Steering Through Chaos Navigating Automated Mobility

Abstract - In recent years, the field of automated mobility has made significant advancements in self-driving vehicle technology. Even so, navigating complex environments such as roads with unexpected pedestrian traffic, uneven surfaces, unforeseen obstacles, and narrow pathways remains challenging. This paper focuses on an innovative AI-based solution to address the unique challenges of autonomous driving in Indian environments. Built on Intel's CARLA, an open-source autonomous driving simulator. Its focus on simulation-based testing ensures the system's reliability in diverse scenarios before real-world deployment, tailored specifically for Indian roads. The system utilizes reinforcement learning for optimal driving strategies with minimal data, computer vision for road signs and obstacle detection, and context-aware navigation to adapt to varying traffic conditions. Offering a promising solution for autonomous driving in regions with similar road conditions by being cost-effective, scalable, and adaptable. Key Words: Automated Vehicles, Simulation-based Software, Computer Vision, Obstacle Detection, Context-aware Navigation, CARLA simulator, Autonomous System Navigation