Research Scientist Intern, Humanoid Control & Dexterous Manipulation (PhD)

Meta•Redmond, WA

2d•$7,650 - $12,667

About The Position

Meta Reality Labs Research is seeking a Research Scientist Intern to help develop the next generation of full-body loco-manipulation and dexterous-hand control policies. The work may span the robotics stack, including data collection and processing, algorithm development and model training, and deploying algorithms on real robots. Interns work with research scientists and engineers on rapid prototyping, iterative research, and debugging algorithms on real hardware. Internships last 12 to 24 weeks, with various start dates throughout the year.

Requirements

Currently has or is pursuing a Ph.D. in robotics, computer graphics, computer vision, machine learning, or a related STEM field.
Experience with physics-based character animation, reinforcement learning and imitation learning, human motion synthesis and retargeting, forward and inverse kinematics and dynamics, and numerical optimization.
Experience collaborating across groups and cultures.
At least 1 year of experience with Python/PyTorch or C/C++, and developing interactive graphics applications.
Must obtain work authorization in the country of employment by the time of hire and maintain it throughout employment.

Nice To Haves

Intent to return to a degree program after completing the internship or co-op.
At least 2 years of experience with physics-based character animation and simulation, and imitation learning.
A record of results demonstrated through accepted papers at leading conference or journal venues, such as RSS, ICRA, IROS, CoRL, SIGGRAPH/Asia/TOG, CVPR, ICCV, ECCV, NeurIPS, ICLR, or ICML.
Experience contributing to open-source projects.
Experience with sim-to-real transfer and deploying simulated policies on real robot hardware.
Experience with perception algorithms such as SLAM and pose estimation.
Experience working with tactile sensors.

Responsibilities

Independently design, develop, and analyze novel research ideas related to robotic simulation and control, focused on loco-manipulation and dexterous manipulation.
Deploy, test, and debug algorithms on real robots for sim-to-real transfer.
Collaborate with researchers and engineers across multiple disciplines.
Regularly communicate the research agenda, progress, and results to diverse audiences.
Publish research results at top conferences and release code that contributes to the robotics field.