Upkies are open-source wheeled biped robots featuring wheels for balance and legs to negotiate uneven terrains. They are designed to be buildable with off-the-shelf tools and components, such as mjbots actuators. You can develop in Python or C++, on Linux or macOS, then deploy your behaviors directly to the robot's Raspberry Pi. There are videos showing the robots in action on the Upkies' YouTube channel.

This repository contains all the materials needed to build and control an Upkie. Questions are welcome in the discussions forum or in the chat room.

Building your own

Upkies come with step by step build instructions.

Getting started

Upkies come out-of-the-box with balancing and locomotion capabilities. You can try them out in simulation using pixi:

pixi run example-follow-joystick

Or using uv:

uv run examples/follow_joystick.py

Once the agent is running, you can direct your Upkie using a gamepad 🎮

Left joystick ⬆️ ⬇️: go forward or backward
Left joystick ⬅️ ➡️: turn left or right
Right button: (B on an Xbox controller, red circle on a PS4 controller) emergency stop 🚨

Click on the robot in the simulator window to apply external forces and see how the robot reacts.

Creating your own behaviors

Software for Upkies comes is packaged in an upkie Python library that you can install from conda or pip:

pip install upkie

This library provides Gymnasium environments to control real and simulation robots with the same code. For instance, the following example balances the robot like a wheeled inverted pendulum in PyBullet:

import gymnasium as gym
import numpy as np
import upkie.envs
 
upkie.envs.register()
 
with gym.make("Upkie-PyBullet-Pendulum", frequency=200.0) as env:
    observation, _ = env.reset()
    gain = np.array([10.0, 1.0, 0.0, 0.1])
    for step in range(1_000_000):
        action = gain.dot(observation).reshape((1,))
        observation, reward, terminated, truncated, _ = env.step(action)
        if terminated or truncated:
            observation, _ = env.reset()

To switch to the real robot, replace "PyBullet" by "Spine" in the environment name.

Examples

There are smaller standalone examples in the examples directory. For instance:

Domain randomization: shows how to add domain-randomization wrappers to an Upkie environment.
Lying genuflection: genuflect while lying on a horizontal floor.
Model predictive control: a self-contained MPC balancer
PD balancer: balance by proportional-derivative feedback to wheel velocities.

Contributing

Contributions are welcome to both the hardware and software of Upkies! Check out the contribution guidelines.

Citation

If you built an Upkie or use parts of this project in your works, please cite the project and its contributors:

@software{upkie,
  title = {{Upkie open source wheeled biped robot}},
  author = {Caron, St\'{e}phane and Perrin-Gilbert, Nicolas and Ledoux, Viviane and G\"{o}kbakan, \"{U}mit Bora and Raverdy, Pierre-Guillaume and Raffin, Antonin and Tordjman{-}{-}Levavasseur, Valentin and Arlaud, Etienne and Duclusaud, Marc},
  url = {https://github.com/upkie/upkie},
  license = {Apache-2.0},
  version = {11.0.0},
  year = {2026}
}

Don't forget to add yourself to the BibTeX above and to CITATION.cff if you contribute to the project.

Building your own

Getting started

Creating your own behaviors

Examples

Contributing

Citation

See also

Reinforcement learning playgrounds for Upkies

Open-source robotics