Unitree Go2 MCP Server: Unleash Natural Language Control with UBOS

The Unitree Go2 MCP Server revolutionizes how you interact with your Unitree Go2 robot. By leveraging the Model Context Protocol (MCP) and integrating seamlessly with the UBOS platform, this server enables users to control the Go2 robot using simple, natural language commands. Forget complex coding; just tell your robot what to do, and watch it happen. The Large Language Model (LLM) interprets your requests, translating them into ROS2 instructions for precise and intuitive control.

Core Functionality: Bridging Language and Action

At its heart, the Unitree Go2 MCP Server acts as a sophisticated translator between human intention and robotic execution. It’s more than just a command interface; it’s a cognitive bridge, allowing the Go2 robot to understand and respond to nuanced instructions. This is achieved through:

• Natural Language Processing (NLP): The LLM at the core of the server understands and interprets human language, extracting the intended action from the command.
• MCP Integration: The Model Context Protocol provides a standardized framework for context exchange, ensuring seamless communication between the LLM and the robot’s control system.
• ROS2 Translation: The interpreted commands are translated into ROS2 instructions, the language understood by the Unitree Go2 robot’s operating system.

Use Cases: Transforming Robotics Across Industries

The Unitree Go2 MCP Server unlocks a wide array of applications across various industries, enabling intuitive robotic control in previously complex scenarios. Consider these transformative use cases:

• Education and Research: Students and researchers can easily experiment with robotic control without deep programming knowledge. Imagine quickly prototyping new robot behaviors or testing different interaction paradigms simply by speaking to the robot.
• Manufacturing and Logistics: Control robots in warehouses or on factory floors with voice commands, streamlining tasks like object retrieval, transportation, and inspection. Picture a supervisor directing a fleet of Go2 robots to manage inventory with spoken instructions.
• Assistance and Support: Enable individuals with mobility limitations to control robots for assistance in daily tasks. Envision a Go2 robot fetching items, navigating obstacles, or providing companionship, all controlled through natural language.
• Entertainment and Events: Create interactive robotic performances or demonstrations where audience members can influence the robot’s actions through spoken commands. Imagine a Go2 robot responding to audience prompts, performing dances, or acting out scenes, all orchestrated by real-time natural language input.
• Search and Rescue Operations: Rapidly deploy and control robots in disaster zones to assess damage, locate survivors, or deliver supplies, all guided by intuitive spoken commands. Envision a team directing a Go2 robot through rubble, using voice commands to navigate and identify potential hazards.

Key Features: Powering Intuitive Robot Control

This server boasts several key features that contribute to its power and ease of use:

• Natural Language Command Interpretation: Control the Unitree Go2 robot using intuitive, natural language commands, eliminating the need for complex programming.
• Seamless ROS2 Integration: Translates natural language commands into ROS2 instructions for direct robot control.
• MCP Compliance: Adheres to the Model Context Protocol for standardized and reliable communication.
• Contextual Understanding: The LLM can interpret commands contextually, understanding the robot’s state and environment for more intelligent responses.
• Easy Installation: Simple setup process, allowing users to quickly get up and running.
• Customizable Functions: Easily extend the robot’s capabilities by adding new functions and behaviors.
• Open Source: Encourages community contributions and customization.

Getting Started: A Step-by-Step Guide

Integrating the Unitree Go2 MCP Server into your workflow is straightforward. Follow these steps to get started:

1. Meet the Requirements:

   • Unitree Go2 robot
   • Ubuntu 20.04 or 22.04
   • ROS2 environment (Humble recommended, or Foxy)

2. Setup unitree_ros2 environment: Follow the instructions at https://github.com/unitreerobotics/unitree_ros2 up to Step 2: Connect and test.

3. Clone the Repository: bash git clone https://github.com/lpigeon/unitree-go2-mcp-server.git cd unitree-go2-mcp-server

4. Install uv (Optional but Recommended): bash curl -LsSf https://astral.sh/uv/install.sh | sh

   or bash pip install uv

   Create a virtual environment and activate it: bash uv venv source .venv/bin/activate

5. Configure MCP Server:

   Set MCP settings in mcp.json. This configuration must be done on the PC connected to the Go2.

   { “mcpServers”: { “unitree-go2-mcp-server”: { “command”: “uv”, “args”: [ “–directory”, “/ABSOLUTE/PATH/TO/PARENT/FOLDER/unitree-go2-mcp-server”, “run”, “server.py” ] } } }

   If using Claude Desktop, find mcp.json using:

   • MacOS: bash code ~/Library/Application Support/Claude/claude_desktop_config.json

   • Linux (Ubuntu): bash code ~/.config/Claude/claude_desktop_config.json

   • Windows: bash code $env:AppDataClaudeclaude_desktop_config.json

6. Set UNITREE_ROS2_SETUP_SH_PATH: Open server.py and update UNITREE_ROS2_SETUP_SH_PATH (e.g., /home/lpigeon/unitree_ros2/setup.sh).

7. Configure Rosbridge (Optional): If using rosbridge, set the LOCAL_IP, ROSBRIDGE_IP, and ROSBRIDGE_PORT in server.py.

8. Verify Go2 Robot Network Connection: Run ros2 topic list in the terminal. You should see the /wirelesscontroller topic. If not, check the network connection.

9. Run Your AI System: Initiate any AI system that has imported unitree-go2-mcp-server.

10. Issue Commands: Start controlling the Go2 robot with natural language commands like, “Make the Go2 robot move forward at a velocity of 0.5 m/s for 3 seconds.”

The Power of UBOS Integration

The Unitree Go2 MCP Server finds a natural and powerful home within the UBOS AI Agent Development Platform. UBOS provides the infrastructure and tools to:

• Orchestrate AI Agents: Seamlessly integrate the Go2 MCP Server with other AI agents to create complex, collaborative robotic systems.
• Connect to Enterprise Data: Grant the Go2 robot access to your enterprise data, enabling data-driven decision-making and actions.
• Build Custom AI Agents: Customize the LLM and functionality of the Go2 MCP Server to meet your specific needs and requirements.
• Develop Multi-Agent Systems: Create sophisticated multi-agent systems where the Go2 robot interacts with other robots and AI agents to achieve common goals.

UBOS empowers you to take the Unitree Go2 MCP Server beyond simple command execution and unlock its full potential as an intelligent, integrated member of your AI ecosystem.

Contributing: Join the Community

We encourage contributions to this project! Whether you’re fixing typos, adding new functions, or suggesting improvements, your help is greatly appreciated. Please follow the contributing guidelines for more details on how to contribute.

With the Unitree Go2 MCP Server and the power of UBOS, the future of robotic control is at your fingertips. Start exploring the possibilities today!