JetBot — NVIDIA Jetson AI Rover

JetBot — NVIDIA Jetson AI Rover

JetBot is an open-source AI robot platform built on the NVIDIA Jetson Nano. It combines a small two-wheeled chassis with a Jetson Nano compute module, a wide-angle camera, and the full NVIDIA JetPack SDK — enabling object following, collision avoidance, and road-following using convolutional neural networks trained with Jupyter notebooks.

JetBot is NVIDIA's official introductory robotics platform, making it one of the most well-documented and widely-built AI robots for beginners and educators.

Hardware

• NVIDIA Jetson Nano Developer Kit (4GB or 2GB)
• Compatible chassis kit (e.g. SparkFun JetBot Kit or Waveshare JetBot Kit)
• Wide-angle USB or CSI camera
• Adafruit Motor FeatherWing or equivalent
• LiPo battery pack
• microSD card (32 GB+)

What It Can Do

• Object following — tracks a sport ball or colored object using a CNN trained in-browser with Jupyter
• Collision avoidance — learns to navigate around obstacles via a head-mounted classifier
• Road following — stays on a track by regressing a steering angle from a front camera
• Teleoperation — browser gamepad control via the interactive widget interface

Source

• Repo: https://github.com/NVIDIA-AI-IOT/jetbot
• License: MIT
• Docs: https://jetbot.org