Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2511.04831 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912692924579840 |
|---|---|
| author | NVIDIA : Mittal, Mayank Roth, Pascal Tigue, James Richard, Antoine Zhang, Octi Du, Peter Serrano-Muñoz, Antonio Yao, Xinjie Zurbrügg, René Rudin, Nikita Wawrzyniak, Lukasz Rakhsha, Milad Denzler, Alain Heiden, Eric Borovicka, Ales Ahmed, Ossama Akinola, Iretiayo Anwar, Abrar Carlson, Mark T. Feng, Ji Yuan Garg, Animesh Gasoto, Renato Gulich, Lionel Guo, Yijie Gussert, M. Hansen, Alex Kulkarni, Mihir Li, Chenran Liu, Wei Makoviychuk, Viktor Malczyk, Grzegorz Mazhar, Hammad Moghani, Masoud Murali, Adithyavairavan Noseworthy, Michael Poddubny, Alexander Ratliff, Nathan Rehberg, Welf Schwarke, Clemens Singh, Ritvik Smith, James Latham Tang, Bingjie Thaker, Ruchik Trepte, Matthew Van Wyk, Karl Yu, Fangzhou Millane, Alex Ramasamy, Vikram Steiner, Remo Subramanian, Sangeeta Volk, Clemens Chen, CY Jawale, Neel Kuruttukulam, Ashwin Varghese Lin, Michael A. Mandlekar, Ajay Patzwaldt, Karsten Welsh, John Zhao, Huihua Anes, Fatima Lafleche, Jean-Francois Moënne-Loccoz, Nicolas Park, Soowan Stepinski, Rob Van Gelder, Dirk Amevor, Chris Carius, Jan Chang, Jumyung Chen, Anka He Ciechomski, Pablo de Heras Daviet, Gilles Mohajerani, Mohammad von Muralt, Julia Reutskyy, Viktor Sauter, Michael Schirm, Simon Shi, Eric L. Terdiman, Pierre Vilella, Kenny Widmer, Tobias Yeoman, Gordon Chen, Tiffany Grizan, Sergey Li, Cathy Li, Lotus Smith, Connor Wiltz, Rafael Alexis, Kostas Chang, Yan Chu, David Fan, Linxi "Jim" Farshidian, Farbod Handa, Ankur Huang, Spencer Hutter, Marco Narang, Yashraj Pouya, Soha Sheng, Shiwei Zhu, Yuke Macklin, Miles Moravanszky, Adam Reist, Philipp Guo, Yunrong Hoeller, David State, Gavriel |
| author_facet | NVIDIA : Mittal, Mayank Roth, Pascal Tigue, James Richard, Antoine Zhang, Octi Du, Peter Serrano-Muñoz, Antonio Yao, Xinjie Zurbrügg, René Rudin, Nikita Wawrzyniak, Lukasz Rakhsha, Milad Denzler, Alain Heiden, Eric Borovicka, Ales Ahmed, Ossama Akinola, Iretiayo Anwar, Abrar Carlson, Mark T. Feng, Ji Yuan Garg, Animesh Gasoto, Renato Gulich, Lionel Guo, Yijie Gussert, M. Hansen, Alex Kulkarni, Mihir Li, Chenran Liu, Wei Makoviychuk, Viktor Malczyk, Grzegorz Mazhar, Hammad Moghani, Masoud Murali, Adithyavairavan Noseworthy, Michael Poddubny, Alexander Ratliff, Nathan Rehberg, Welf Schwarke, Clemens Singh, Ritvik Smith, James Latham Tang, Bingjie Thaker, Ruchik Trepte, Matthew Van Wyk, Karl Yu, Fangzhou Millane, Alex Ramasamy, Vikram Steiner, Remo Subramanian, Sangeeta Volk, Clemens Chen, CY Jawale, Neel Kuruttukulam, Ashwin Varghese Lin, Michael A. Mandlekar, Ajay Patzwaldt, Karsten Welsh, John Zhao, Huihua Anes, Fatima Lafleche, Jean-Francois Moënne-Loccoz, Nicolas Park, Soowan Stepinski, Rob Van Gelder, Dirk Amevor, Chris Carius, Jan Chang, Jumyung Chen, Anka He Ciechomski, Pablo de Heras Daviet, Gilles Mohajerani, Mohammad von Muralt, Julia Reutskyy, Viktor Sauter, Michael Schirm, Simon Shi, Eric L. Terdiman, Pierre Vilella, Kenny Widmer, Tobias Yeoman, Gordon Chen, Tiffany Grizan, Sergey Li, Cathy Li, Lotus Smith, Connor Wiltz, Rafael Alexis, Kostas Chang, Yan Chu, David Fan, Linxi "Jim" Farshidian, Farbod Handa, Ankur Huang, Spencer Hutter, Marco Narang, Yashraj Pouya, Soha Sheng, Shiwei Zhu, Yuke Macklin, Miles Moravanszky, Adam Reist, Philipp Guo, Yunrong Hoeller, David State, Gavriel |
| contents | We present Isaac Lab, the natural successor to Isaac Gym, which extends the paradigm of GPU-native robotics simulation into the era of large-scale multi-modal learning. Isaac Lab combines high-fidelity GPU parallel physics, photorealistic rendering, and a modular, composable architecture for designing environments and training robot policies. Beyond physics and rendering, the framework integrates actuator models, multi-frequency sensor simulation, data collection pipelines, and domain randomization tools, unifying best practices for reinforcement and imitation learning at scale within a single extensible platform. We highlight its application to a diverse set of challenges, including whole-body control, cross-embodiment mobility, contact-rich and dexterous manipulation, and the integration of human demonstrations for skill acquisition. Finally, we discuss upcoming integration with the differentiable, GPU-accelerated Newton physics engine, which promises new opportunities for scalable, data-efficient, and gradient-based approaches to robot learning. We believe Isaac Lab's combination of advanced simulation capabilities, rich sensing, and data-center scale execution will help unlock the next generation of breakthroughs in robotics research. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_04831 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning NVIDIA : Mittal, Mayank Roth, Pascal Tigue, James Richard, Antoine Zhang, Octi Du, Peter Serrano-Muñoz, Antonio Yao, Xinjie Zurbrügg, René Rudin, Nikita Wawrzyniak, Lukasz Rakhsha, Milad Denzler, Alain Heiden, Eric Borovicka, Ales Ahmed, Ossama Akinola, Iretiayo Anwar, Abrar Carlson, Mark T. Feng, Ji Yuan Garg, Animesh Gasoto, Renato Gulich, Lionel Guo, Yijie Gussert, M. Hansen, Alex Kulkarni, Mihir Li, Chenran Liu, Wei Makoviychuk, Viktor Malczyk, Grzegorz Mazhar, Hammad Moghani, Masoud Murali, Adithyavairavan Noseworthy, Michael Poddubny, Alexander Ratliff, Nathan Rehberg, Welf Schwarke, Clemens Singh, Ritvik Smith, James Latham Tang, Bingjie Thaker, Ruchik Trepte, Matthew Van Wyk, Karl Yu, Fangzhou Millane, Alex Ramasamy, Vikram Steiner, Remo Subramanian, Sangeeta Volk, Clemens Chen, CY Jawale, Neel Kuruttukulam, Ashwin Varghese Lin, Michael A. Mandlekar, Ajay Patzwaldt, Karsten Welsh, John Zhao, Huihua Anes, Fatima Lafleche, Jean-Francois Moënne-Loccoz, Nicolas Park, Soowan Stepinski, Rob Van Gelder, Dirk Amevor, Chris Carius, Jan Chang, Jumyung Chen, Anka He Ciechomski, Pablo de Heras Daviet, Gilles Mohajerani, Mohammad von Muralt, Julia Reutskyy, Viktor Sauter, Michael Schirm, Simon Shi, Eric L. Terdiman, Pierre Vilella, Kenny Widmer, Tobias Yeoman, Gordon Chen, Tiffany Grizan, Sergey Li, Cathy Li, Lotus Smith, Connor Wiltz, Rafael Alexis, Kostas Chang, Yan Chu, David Fan, Linxi "Jim" Farshidian, Farbod Handa, Ankur Huang, Spencer Hutter, Marco Narang, Yashraj Pouya, Soha Sheng, Shiwei Zhu, Yuke Macklin, Miles Moravanszky, Adam Reist, Philipp Guo, Yunrong Hoeller, David State, Gavriel Robotics Artificial Intelligence We present Isaac Lab, the natural successor to Isaac Gym, which extends the paradigm of GPU-native robotics simulation into the era of large-scale multi-modal learning. Isaac Lab combines high-fidelity GPU parallel physics, photorealistic rendering, and a modular, composable architecture for designing environments and training robot policies. Beyond physics and rendering, the framework integrates actuator models, multi-frequency sensor simulation, data collection pipelines, and domain randomization tools, unifying best practices for reinforcement and imitation learning at scale within a single extensible platform. We highlight its application to a diverse set of challenges, including whole-body control, cross-embodiment mobility, contact-rich and dexterous manipulation, and the integration of human demonstrations for skill acquisition. Finally, we discuss upcoming integration with the differentiable, GPU-accelerated Newton physics engine, which promises new opportunities for scalable, data-efficient, and gradient-based approaches to robot learning. We believe Isaac Lab's combination of advanced simulation capabilities, rich sensing, and data-center scale execution will help unlock the next generation of breakthroughs in robotics research. |
| title | Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning |
| topic | Robotics Artificial Intelligence |
| url | https://arxiv.org/abs/2511.04831 |