PhD Robotics AI Research Intern: Real-World Dexterity

Roboticsconnected; Fremont (CA); On-site; USD 40,000 - 60,000; Full time

Job summary

DEXMATE is seeking PhD research interns to advance robot learning on real hardware. You will develop new algorithms for dexterous manipulation, navigation, and control, and validate ideas on Dexmate robots with our research and engineering teams.

You will train and fine-tune large models, conduct multi-disciplinary research (robotics, RL/IL, perception, ML/DL), and design learning-based control systems using real robots with our team in California.

Qualifications

• PhD candidate in CS/Robotics/EE/ME or related field.
• 2+ years of independent research in robotics or AI.
• Experience with PyTorch and deep learning.
• Authorized to work in the U.S. during employment.
• Strong analytical, problem-solving, and communication skills.
• Publications in top conferences/journals preferred.

Responsibilities

• Develop algorithms for dexterous manipulation, navigation, and control.
• Train and fine-tune large models and validate on Dexmate robots.
• Collaborate with research and engineering teams.
• Coordinate experiments and analyze results on real hardware.
• Publish findings at conferences/journals when possible.

Skills

PhD in CS/Robotics/EE/ME

Robotics passion

Embodied AI research

PyTorch

US work authorization

Analytical & communication

Independent research (2+ yrs)

Robot learning (RL/IL)

Robot simulators (Isaac Gym/Sim, SAPI,

ML experiment setup and analysis

ML/DL systems building

Education

PhD candidate in CS/Robotics/EE/ME

Tools

PyTorch

Robot simulators

Job description

DEXMATE is seeking PhD research interns to advance robot learning on real hardware. You will develop new algorithms for dexterous manipulation, navigation, and control, and validate ideas on Dexmate robots with our research and engineering teams.

You will train and fine-tune large models, conduct multi-disciplinary research (robotics, RL/IL, perception, ML/DL), and design learning-based control systems using real robots with our team in California.