Robot learning frameworks

Research

Reinforcement-learning tools for robotic locomotion training.

Research

Isaac Gym environments for legged robot learning.

Research

StarVLA provides modular code for developing vision-language-action models.

Research

EmbodiInfer provides inference and reinforcement-learning rollout infrastructure for embodied models.

Research

EmbodiRun provides a runtime connecting embodied models with compute nodes and robots or simulators.

Research

Dexbotic provides a toolbox for vision-language-action model development.

Research

ROS-Industrial support packages for Universal Robots arms.

Research

URDFs and assets for robot manipulation experiments.

Research

SAPIEN simulation platform for embodied AI.

Research

An open ecosystem for robot learning.

Research

Toolkit for model-based robotics design, simulation, and analysis.

Research

ROS 2 core platform and development resources.

Research

ROS-Industrial support packages for Fanuc robots.

Research

SOFA-based soft-robot RL with domain randomization.

Research

ROS packages for Hello Robot Stretch mobile manipulators.

Research

RLinf ActProbe adds probe-based early termination to embodied-policy reinforcement learning.

Research

A plug-and-play deploy framework for robots. Just deploy, just do.

Research

RoboHive provides a framework and task suites for robot learning.

Research

ROS 2 driver for the Kinova Gen3 robot arm.

Research

Robot model collection for OpenRAVE.

Research

A visual fiducial system for robotics research.

Research

Main software repository for the iCub humanoid robot.

Research

Examples for deploying ROS and ROS 2 workloads on Kubernetes and KubeEdge.

Research

Research code for injecting robotic affordances into multimodal language models.