JetRacer provides an autonomous AI racecar platform built around the NVIDIA Jetson Nano, targeting hobbyists and researchers interested in high-speed AI applications. It enables users to build and program AI racecars interactively via web browser, optimizing for fast AI pipelines and pushing performance boundaries.

How It Works

The project utilizes Jupyter Notebooks for interactive programming and AI model training. It emphasizes optimizing models with NVIDIA TensorRT for high frame rates and real-time performance, facilitating the development of applications like road following.

Quick Start & Requirements

• Install/Run: Follow hardware and software setup guides linked in the README.
• Prerequisites: NVIDIA Jetson Nano, specific parts from provided bills of materials (Latrax Rally or Tamiya TT02 chassis), JetPack 4.5.1 based image.
• Setup: Requires ordering parts, hardware assembly, and software configuration. Links to detailed setup guides are available.

Highlighted Details

• Offers two chassis options: Latrax Rally (1/18th scale, moderate speed, ~$400) and Tamiya TT02 (1/10th scale, high speed, ~$600).
• Examples include basic motion control and AI-powered road following with interactive training.
• Focuses on optimizing AI models using NVIDIA TensorRT for high framerates.

Maintenance & Community

• Users can ask questions by creating issues on the GitHub repository.
• Related projects include JetBot, JetCam, JetCard, and torch2trt.

Licensing & Compatibility

• License details are not explicitly stated in the README.

Limitations & Caveats

• The project is based on JetPack 4.5.1, which may be outdated. Specific hardware components are required, and build costs vary.