About the job
The Isaac Loco-Manipulation team develops innovative robot learning technologies that enable humanoid and mobile robots for complex dexterous manipulation and loco-manipulation tasks. This internship combines ambitious applied research and thorough engineering across GR00T and Cosmos foundation models, Isaac Lab, Newton, synthetic data generation, simulation, and physical robot testing. As an intern, you will own a focused applied research project and collaborate with world-class researchers and engineers to turn novel ideas into validated robot capabilities. This outstanding opportunity allows you to immerse yourself in the forefront of AI and robotics, making a lasting impact on the future of technology.
Responsibilities
Develop novel robot learning methods for advancing performance in humanoids and mobile manipulators for whole-body dexterous tasks.
Co-develop and advance GR00T and Cosmos robotics foundation models.
Develop robot-learning technologies to enable large scale deployment of robot autonomy.
Lead an applied research project from idea conception, prototype development, publication, and software open-source contribution.
Qualifications
Minimum
Pursuing a PhD degree or Master Degree in Robotics, Computer Science, or a related field.
Proven strong robotics research or engineering experience through academic projects, publications, and real robot experience.
Proficiency in Python and C++, with hands-on experience using deep-learning frameworks such as PyTorch, JAX, or TensorFlow.
Experience with physics simulators like Isaac Sim, Isaac Lab, or MuJoCo and transferring data or behaviors between simulated and physical systems.
Experience with world modeling and world action models.
Project or research experience in robot learning, including imitation learning or reinforcement learning, and robotics foundation models.
Preferred
Strong publication record in leading robotics, computer-vision, or machine-learning journals or conferences such as CoRL, RSS, Science Robotics, ICRA, and CVPR.
Experience in learning robot behaviors from human videos or demonstrations.
Research or project depth in dexterous bimanual manipulation, whole-body control, or humanoid robotics.
Passion for transforming research prototypes into scalable robotics products widely embraced by the physical AI community. Successfully implement your ideas into impactful technologies!