Research Scientist Intern, Robotics

Meta•Burlingame, CA

2d•$7,650 - $12,667

About The Position

Meta is seeking a Research Scientist Intern to advance robot learning and embodied intelligence by building systems that perceive, reason, and act in the physical world. The work combines research and problem-solving with hands-on experimentation using real robot hardware and simulation to design, develop, and evaluate novel policies, architectures, and prototypes. Interns join a focused incubation team and collaborate with scientists and engineers across the organization. Internships last 12 to 24 weeks, with various start dates throughout the year.

Requirements

Currently hold or be in the process of obtaining a Ph.D. in Robotics, Computer Science, Computer Vision, Artificial Intelligence, Electrical or Mechanical Engineering, Controls, or a relevant technical field.
Experience with Python, C++, C, Java, or another related programming language.
Experience building systems based on machine learning and/or deep learning methods.
Obtain work authorization in the country of employment by the time of hire and maintain ongoing work authorization during employment.

Nice To Haves

A track record of significant results demonstrated by grants, fellowships, patents, or first-authored publications at leading workshops or conferences such as RSS, CoRL, ICRA, IROS, NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, or similar.
Experience working and communicating cross-functionally in a team environment.
Experience advancing AI techniques for robotics or computer vision, including core contributions to open-source robotics or machine-learning libraries and frameworks such as ROS/ROS 2, MuJoCo, Isaac Sim/Lab, Drake, PyBullet, or PyTorch.
Experience deploying learned policies or perception systems on real robot hardware, such as manipulators, mobile robots, humanoids, or dexterous hands.
Experience with simulation, domain randomization, and sim-to-real transfer.
Experience with classical robotics foundations, including state estimation, SLAM, calibration, motion planning, and optimal or model-predictive control.
Publications or experience in robot learning, reinforcement learning, machine learning, AI, computer vision, optimization, computer science, statistics, applied mathematics, or data science.
Experience solving analytical problems using quantitative approaches.
Experience setting up machine-learning or robot experiments and analyzing results, including real-world evaluation protocols.
Experience manipulating and analyzing complex, large-scale, high-dimensional data from varying sources and sensor modalities.
Experience using theoretical and empirical research to solve problems.
Experience with deep-learning frameworks.
Intent to return to a degree program after completing the internship or co-op.

Responsibilities

Develop state-of-the-art algorithms for robot perception, manipulation, navigation, and control using modern deep learning and foundation-model techniques.
Train and evaluate robot learning policies, including imitation learning, reinforcement learning, and vision-language-action models, in simulation and on physical hardware.
Help improve the efficiency, robustness, scalability, and real-time performance of algorithms deployed on robot platforms.
Conduct research to advance the science and technology of intelligent machines.
Conduct research enabling the learning of semantics from multimodal and embodied data, including images, video, depth, point clouds, tactile and force signals, proprioception, language, and audio.
Contribute to data collection, teleoperation, and sim-to-real transfer pipelines that support large-scale robot learning.
Collaborate with researchers and cross-functional partners across hardware, controls, software, and product, and communicate research plans, progress, and results.
Publish research results and contribute to research applicable to Meta product development.