Skip to content

Software

Open software that robots run on, with the license read from each project's license file, whether it is still maintained, and the robots on Robotheca that use it. A reference, not a download site: every entry links to its source.

Filters

Category
Difficulty

How hard it is to get running, chosen by a person at Robotheca for 276 of 276 software.

Options

Commercial use: its licenses together (code and assets) allow it, as Robotheca reads them. Maintained: a commit in the last 12 months, not archived or discontinued.

Clear filters

36 projects of 276

Page 2 of 2 · newest commit first

    • Software
    • Humanoids and robot learning

    A ROS 2 package set that lets developers talk to Unitree Go2, B2, and H1 robots using standard ROS 2 messages over CycloneDDS, without wrapping the SDK.

    Code license
    BSD-3-Clause
    Commercial use
    Yes
    Difficulty
    Intermediate
    Status
    Maintained
    • Software
    • Humanoids and robot learning

    UMI is a handheld gripper and software pipeline for collecting robot manipulation demonstrations and training visuomotor policies from them, deployable on UR5 and ARX X5 arms.

    Code license
    MIT
    Commercial use
    Yes
    Difficulty
    Advanced
    Status
    Maintained
    • Software
    • Humanoids and robot learning

    ROSA is an AI assistant built on LangChain that lets people interact with ROS 1 and ROS 2 robotic systems through natural language queries.

    Code license
    Apache-2.0
    Commercial use
    Yes
    Difficulty
    Intermediate
    Status
    Maintained
    • Software
    • Humanoids and robot learning

    A C++ toolbox for formulating and solving nonlinear optimal control problems, aimed at real-time Model Predictive Control (MPC) for robotics.

    Code license
    BSD-3-Clause
    Commercial use
    Yes
    Difficulty
    Advanced
    Status
    Maintained
    • Software
    • Humanoids and robot learning

    A reinforcement learning toolkit for training and deploying motion control policies on Unitree robots such as Go2, H1, H1_2 and G1. It covers training, playback, simulation and real-robot deployment.

    • Closed core
    • Stale
    Code license
    BSD-3-Clause
    Commercial use
    Unknown
    Difficulty
    Advanced
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A training environment built on NVIDIA Isaac Gym for teaching legged robots, such as ANYmal, to walk over rough terrain with sim-to-real transfer features.

    • Closed core
    • Stale
    • Moved
    Code license
    BSD-3-Clause
    Commercial use
    Unknown
    Difficulty
    Advanced
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A C++ motion control framework for the Qinglong humanoid robot, combining MPC and WBC, runnable in the MuJoCo simulator with walking, jumping and blind obstacle stepping examples.

    • Stale
    Code license
    Apache-2.0
    Commercial use
    Yes
    Difficulty
    Advanced
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A model-based control stack for legged robots combining nonlinear MPC and whole-body control, built on OCS2 and ros-control for ROS. It is no longer supported by its authors.

    • Stale
    Code license
    BSD-3-Clause
    Commercial use
    Yes
    Difficulty
    Advanced
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A learning framework that trains robot visuomotor policies by generating actions with a diffusion model, with reproduced benchmarks in simulation and a real UR5 robot setup.

    • Stale
    Code license
    MIT
    Commercial use
    Yes
    Difficulty
    Intermediate
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A ROS software framework for building and controlling quadruped robots, with a gait controller, simulation, and navigation support. It targets both new robot designs and developing control algorithms.

    • Stale
    Code license
    BSD-3-Clause
    Commercial use
    Yes
    Difficulty
    Intermediate
    Status
    Stale
    • Software
    • Humanoids and robot learning

    ACT is an imitation-learning policy that predicts action chunks with transformers, with simulated tasks for training and evaluating it in sim or on real robots.

    • Stale
    Code license
    MIT
    Commercial use
    Yes
    Difficulty
    Intermediate
    Status
    Stale
    • Software
    • Humanoids and robot learning

    A robot and simulation software project for MIT Biomimetics' Cheetah quadruped robots, with a shared dynamics library, robot control code, and a Qt-based simulator.

    • Stale
    Code license
    MIT
    Commercial use
    Yes
    Difficulty
    Advanced
    Status
    Stale

Projects 25–36 of 36

How Robotheca reads each entry

  • The code license comes from the project's license files at a pinned commit, never from a badge.
  • Maintenance comes from the newest commit: active within 90 days, maintained within 12 months, stale after that.
  • Nothing is hosted or run here: every page links to the project itself.