Welcome to RLRoverLab! This project implements a suite of Reinforcement Learning (RL) agents using Isaac Sim and Isaac Lab. Our suite currently supports a variety of tasks within navigation and manipulation, with ongoing efforts to expand our offerings.
- Navigation and Manipulation Tasks: Implementations of RL agents designed for navigation and manipulation tasks, we are working on integrating more tasks.
- Isaac Sim and Isaac Lab Integration: Utilizes the advanced simulation environments of Isaac Sim and the Isaac Lab framework for realistic task scenarios.
- Expandable Framework: Architecture designed for easy extension with new tasks and functionalities.
To get started with RLRoverLab, please refer to our Installation Guide. The guide provides comprehensive steps for setting up the suite using Docker as well as instructions for native installation.
Parralel.mp4
Parralel_video4_1440p.mp4
rover_video.mp4
If you have questions, suggestions, feel free to contact us.
- Contact Us: For direct inquiries, reach out to Anton Bjørndahl Mortensen at antonbm2008@gmail.com.
Please cite this paper if you use this suite in your work:
@inproceedings{mortensen2024rlroverlab,
title={RLRoverLAB: An Advanced Reinforcement Learning Suite for Planetary Rover Simulation and Training},
author={Mortensen, Anton Bj{\o}rndahl and B{\o}gh, Simon},
booktitle={2024 International Conference on Space Robotics (iSpaRo)},
pages={273--277},
year={2024},
organization={IEEE}
}