About the Role
We are looking for a graduate-level Robot Learning Engineering Intern to join our Skills team and support the development of learned manipulation behaviors for our humanoid robots. You will work at the intersection of robot learning, teleoperation, data collection, and on-robot evaluation, helping the team build and validate learning workflows that can support future deployment on humanoid systems.
This internship is especially well-suited for a graduate student who wants hands-on experience with a self-contained robot learning testbed and is excited to explore how high-quality demonstrations, sensing, and evaluation can be used to develop force-aware or contact-rich manipulation policies. You will work closely with engineers developing Learning from Demonstration (LfD) systems and learned robot skills, contributing to the tooling, experiments, and analysis needed to make these approaches useful in practice.
Key Responsibilities
- Support demonstration data collection for learned robot behaviors using teleoperation and other operator-in-the-loop systems
- Help build and improve a robot learning testbed, including integration of teleop interfaces, cameras, and other sensing required for data collection and evaluation
- Assist with experiments focused on contact-rich or force-aware manipulation behaviors
- Build and improve tooling for data ingestion, annotation, validation, replay, and analysis
- Assist in evaluating learned policies in simulation and on real robot hardware
- Partner with engineers to debug failures and improve system robustness
- Analyze experiments and generate insights around data quality, policy behavior, and deployment readiness
- Contribute to internal tools and infrastructure for robot learning development
- Document findings and communicate results to mentors and cross-functional stakeholders
About You
- Currently pursuing an MS or PhD in Robotics, Computer Science, Machine Learning, or a related field
- Strong software engineering fundamentals and proficiency in Python
- Background in robotics, machine learning, embodied AI, controls, or autonomous systems
- Familiarity with at least some of the following: imitation learning, reinforcement learning, robot manipulation, force control, teleoperation, robot sensors, or object perception
- Strong analytical, experimental, and debugging skills
- Comfortable working hands-on with robotic systems in a lab environment
Bonus Qualifications
- Experience with behavior cloning, Learning from Demonstration, offline RL, or learned control policies
- Experience with robot data collection systems, VR/XR tools, haptics, or teleoperation interfaces
- Experience with force/torque sensing or contact-rich manipulation tasks
- Experience evaluating algorithms on real hardware rather than simulation alone
This is an onsite internship at our office in Pittsburgh, PA - minimum 4 days/week onsite.
Duration: Internship (3-6+ months, flexible)