ScenarIO and gym-gz

ScenarIO and gym-gz#

This project targets both control and robot learning research domains:

  • Researchers in robotics and control can simulate their robots with familiar tools like Gazebo and URDF/SDF, without the need to rely on any middleware.

  • Researchers in robot learning can quickly develop new robotic environments that can scale to hundreds of parallel instances.

We provide two related subprojects to each of these categories:

  1. ScenarIO provides APIs to interface with the robots.

  2. gym-gz helps structuring environments compatible with OpenAI Gym, while minimizing boilerplate code and providing common rigid-body dynamics utilities.

Check the sections What is ScenarIO and What is gym-gz for more details, and visit Motivations for an extended overview.

For a quick practical introduction, visit the Getting Started page.

If you use this project for your research, please check the FAQ about how to give credit.

pendulum_swing

panda_grasping

icub_stepping