PetoiCamp/OpenCat-Quadruped-Robot

What it solves

OpenCat provides an accessible, open-source framework for building and programming agile quadruped robots. It removes the complexity of low-level hardware management—such as gait coordination, servo control, and IMU integration—allowing researchers, educators, and makers to focus on higher-level behaviors and AI applications without needing a massive budget or a specialized lab.

How it works

The framework is built on Arduino and Raspberry Pi, supporting multiple languages including C/C++, Python, and block-based coding. It coordinates servos for movement (walking, running, jumping) and integrates sensors like cameras, touch, and IMUs. For advanced users, it is ROS compatible, enabling the use of SLAM, navigation, and perception pipelines. It also supports "sim-to-real" workflows, where reinforcement learning models can be tested in simulation and then deployed to the physical robot.

Who it’s for

• Researchers and Academics: Those conducting embodied robotics experiments or gait research.
• Educators: K-12 teachers and university professors teaching robotics programming and physical AI.
• Makers: Hobbyists building custom robotic pets or participating in hackathons.

Highlights

• Open Ecosystem: Open hardware and software with no vendor lock-in, supporting 3D-printable shells and mounts.
• AI Ready: Compatible with NVIDIA Isaac simulations, Tiny ML for imitation learning, and AI co-processors like Jetson Nano.
• Broad Hardware Support: Works across multiple robot platforms (Bittle, Nybble, and the upcoming Quaddle).
• Extensible: Supports integration with external APIs (e.g., ChatGPT, Azure AI) and IoT fleet management via AWS.