AI Robotics Researcher Intern (Dexterous Manipulation)
Core
Researching embodied AI and dexterous manipulation by translating human-object interaction data into executable robotic behaviors for physical world intelligence.
Role type
AI Robotics Research Intern (Dexterous Manipulation)
Builds
Robust pipelines for converting human multi-modal data into actionable robot motor skills and sim-to-real validation of dexterous manipulation policies.
Domain
Robotics, Embodied AI, Machine Learning
Deliverable
production ML models | research
Required skills
robot learning and control, reinforcement learning, imitation learning, world modeling, representation learning, Generative Models, Transformer architectures, robotic manipulation systems, Python, PyTorch, JAX, TensorFlow, ROS/ROS2, MuJoCo, Isaac Sim, PyBullet
Preferred skills
dexterous manipulation, multi-finger robotic hands, in-hand manipulation, grasp optimization, sim-to-real transfer, contact modeling, tactile sensing, force/torque feedback, 3D perception, geometric representations, reinforcement learning in continuous control, model-based methods, real-time policy execution
Technologies
PyTorch, JAX, TensorFlow, ROS, ROS2, MuJoCo, Isaac Sim, PyBullet
Responsibilities
Develop scalable frameworks for transferring human-object interaction skills to robotic embodiments, architect autonomous pipelines for processing visual and human glove-collected data, implement generative architectures for synthesizing physically grounded trajectories, explore cross-embodiment representations for joint training, research distillation and retargeting techniques for sim-to-real deployment, utilize vision-language foundation models for semantic scene understanding
Seniority
Intern
