Research Intern Data Collection & Deployment - Teleoperation, Data Pipelines, On-Robot Systems (MS/PhD, 612 months)

Join our team to help build the data engine behind general-purpose robot policies. You'll work on the pipeline from teleoperated demonstration collection through curation and quality control to on-robot deployment — the systems that turn robot time into training data, and trained policies into robots that work in the real world. You'll own a concrete piece of this stack end to end, working directly with our engineering and research team on real hardware. We care more about your ability to build things that work on real robots than about any particular robot, task, or sensor you've used before.

Requirements: Currently pursuing an MS or PhD in Robotics, Computer Science, Electrical Engineering, or related field
Strong software engineering skills in Python (C++ a plus) in Linux environments: you write clean, working code and can debug across the stack
Hands-on experience with real robots or physical systems — through research, projects, competitions, or prior internships
Familiarity with the robot data workflow: collecting demonstrations (teleoperation or kinesthetic), logging multimodal sensor streams, and organizing data for training
Comfortable working with sensor data (RGB/depth cameras, proprioception, tactile, force-torque) — capture, synchronization, and visualization
Rigorous and self-directed: you can take a loosely specified problem, break it down, and drive it to a working result in a limited time
(+) Experience with ROS 2 or comparable robotics middleware, or real-time systems
(+) Experience deploying or evaluating learned policies on real hardware
(+) Familiarity with robot learning (imitation learning, vision-language-action models) — enough to understand what models need from data

• (+) Experience with dexterous hands, tactile sensing, or contact-rich manipulation setups

• Paid internship with competitive compensation

• Work on cutting-edge problems in robot learning and manipulation

• Mentorship from researchers and engineers working at the frontier of embodied intelligence

• Access to real robot hardware and large-scale robot datasets

• Opportunity to publish or contribute to high-impact research alongside product-driven development

• Build and improve components of the teleoperation and demonstration-collection stack — rigs, operator interfaces, and collection workflows

• Support data collection operations on real robots and help raise throughput and data quality

• Contribute to the pipeline from robot to training set: ingestion, time synchronization of multimodal sensor streams, curation, filtering, and annotation tooling

• Build QA, metrics, and visualization tooling that keeps datasets consistent and trustworthy

• Help deploy trained policies to real hardware and run on-robot evaluations

• Assist with hardware and sensor bring-up — cameras, tactile, force-torque — on collection and deployment rigs

• Work directly with the AI, hardware, and perception teams, and own a scoped project from idea to a working, demonstrated result