MINT-SJTU/RoboClaw

RoboClaw is an Embodied AI Assistant.

What it solves

RoboClaw provides an open-source framework for building embodied intelligence assistants, streamlining the process of moving from robot arm setup and calibration to data collection, training, and inference.

How it works

It operates as an embodied framework that uses domain contracts and an assembly-centered onboarding controller to manage the workflow. The system supports conversational arm setup, teleoperation for data collection, and integration with ROS2 for the execution layer.

Who it’s for

This project is designed for developers and researchers in embodied AI who need a structured environment for developing, training, and simulating robot arm capabilities.

Highlights

  • Open-source embodied intelligence assistant.
  • Supports the full workflow: setup, calibration, teleoperation, data collection, training, and inference.
  • Integrates with ROS2 and supports simulator adaptation.
  • Includes a web dashboard for managing the embodied workflow.

Related

  • Project
    openclaw/openclaw

    OpenClaw is a self-hosted personal AI assistant that connects AI models and tools to your favorite messaging channels and local devices.

  • Project
    Seeed-Projects/reBot-DevArm

    An open-source robotic arm project providing a full hardware and software stack to lower the barrier for learning and deploying Embodied AI.

  • Project
    PetoiCamp/OpenCat-Quadruped-Robot

    An open-source framework for quadruped robots that handles gait coordination and servo control, enabling the development of AI-driven robotic pets and embodied intelligence research.

  • Project
    enactic/openarm

    An open-source 7DOF humanoid arm and standardized environment designed for physical AI research, imitation learning, and safe human-robot interaction.

  • Project
    aiming-lab/MetaClaw

    An agent proxy that enables AI assistants to meta-learn and evolve through real-world conversations using skill injection and asynchronous RL fine-tuning.