F1TENTH Autonomous Racing

Project Overview

Led the development of a 1/10 scale autonomous racing vehicle for F1TENTH competition, implementing advanced robotics systems on both hardware and software levels. The project showcased expertise in autonomous navigation, real-time control systems, and advanced simulation environments.

Technical Implementation

• Hardware Integration: Developed a complete autonomous platform using Raspberry Pi 4B as the main computer, integrated with VESC motor controller and LIDAR sensor. The system architecture focused on low-latency performance and reliable sensor fusion for racing conditions.
• Software Architecture: Created an advanced ROS2-based navigation stack featuring custom nodes for LIDAR processing, PID-based wall following, and gap detection algorithms. Implemented Nav2 framework integration for mapping and autonomous navigation, with emphasis on real-time performance optimization.
• Development Environment: Established a robust development pipeline using Docker containers, the default F1TENTH simulator and NVIDIA Isaac Sim for testing.

Key Achievements

• Achieved high-performance autonomous navigation with sub-30ms latency.
• Successfully implemented advanced control algorithms for racing scenarios.
• Developed comprehensive simulation and testing framework.