Orbit: A Unified Simulation Framework for Interactive Robot Learning Environments

• DOI: 10.1109/lra.2023.3270034
• Publication Date: 2023-04-25T18:43:59Z
• Authors: Mayank Mittal; Calvin Yu; Qinxi Yu; Jingzhou Liu; Nikita Rudin; David Hoeller; Jia Lin Yuan; Ritvik Singh; Yunrong Guo; Hammad Mazhar; Ajay Mandlekar; Buck Babich; Gavriel State; Marco Hutter; Animesh Garg
• Keywords: Benchmark (surveying)

We present <sc xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Orbit</small> , a unified and modular framework for robot learning powered by xmlns:xlink="http://www.w3.org/1999/xlink">Nvidia</small> Isaac Sim. It offers design to easily efficiently create robotic environments with photo-realistic scenes high-fidelity rigid deformable body simulation. With we provide suite of benchmark tasks varying difficulty– from single-stage cabinet opening cloth folding multi-stage such as room reorganization. To support working diverse observations action spaces, include fixed-arm mobile manipulators different physically-based sensors motion generators. allows training reinforcement policies collecting large demonstration datasets hand-crafted or expert solutions in matter minutes leveraging GPU-based parallelization. In summary, offer an open-sourced that readily comes 16 platforms, 4 sensor modalities, 10 generators, more than 20 tasks, wrappers libraries. this framework, aim various research areas, including representation learning, imitation task planning. hope it helps establish interdisciplinary collaborations these communities, its modularity makes extensible applications the future.