pollen-robotics/microduck

A Tiny biped duck robot 🦆

What it solves

Microduck provides the "brain" and control software for a small biped robot, enabling it to perform complex physical movements like walking, rolling, picking up objects, and standing back up after being knocked over.

How it works

The system runs on a Rockchip RK3566 and uses a 50 Hz control loop to drive fifteen servos based on neural policies. These policies are trained using reinforcement learning (MuJoCo and PPO) in a separate repository and exported to ONNX format for the robot to load and execute. The software architecture consists of several Rust-based daemons (such as robotd for control and updaterd for updates) that communicate via JSON-RPC over Unix sockets.

Who it’s for

It is designed for robot owners, developers building on the platform, and researchers interested in sim-to-real reinforcement learning for bipedal locomotion.

Highlights

  • RL-Driven Movement: Uses neural policies trained via PPO and MuJoCo for natural movement.
  • Multi-modal Locomotion: Capable of walking, rolling on wheels, and performing specific actions like kicking a ball.
  • Robust Update System: Features a verified, health-gated, and reversible update mechanism to prevent bricking.
  • Simulation Support: Includes a simulated duck environment using MuJoCo to test daemons without physical hardware.

Related

  • Project
    AI-FanGe/Microduck-build-tutorial

    A compact biped robot project providing a full pipeline from MuJoCo-based RL training to real-world deployment on a Raspberry Pi Zero 2 W using ONNX policies.

  • Dispatch
    Pollen Robotics Microduck: An Open-Source Biped for RL Training

    Pollen Robotics has launched Microduck, a 25cm open-source biped robot designed for reinforcement learning (RL) training via a sim-to-real pipeline using MuJoCo.

  • Project
    pollen-robotics/microduck_rl

    RL training environments for the Microduck bipedal robot that use high-fidelity actuator modeling and domain randomization to enable sim2real transfer of complex gaits and tricks.

  • Project
    fanhao375/microduck-replica

    A reverse-engineering project that reconstructs the assembly drawings, CAD models, and hardware specifications of the Microduck bipedal robot from its official simulation files.

  • Project
    HansZ8/RoboJuDo

    A plug-and-play deployment framework for humanoid robots that decouples controllers, environments, and policies to simplify the deployment of AI control models on real hardware and in simulation.