Robotics Engineering Intern: Data and Evaluation at Origami Robotics

Origami Robotics is building our own robotic hand and training physical intelligence models for dexterous manipulation. We are building the full loop ourselves: hardware, teleoperation, data collection, evaluation, and learned policies.

About the role

This internship is for someone who wants to work close to the robot. You will build tools and workflows that help us collect better data, evaluate robot behavior, debug failures, and improve the system through real physical experiments.

Responsibilities

• Help build and improve robot data collection workflows on physical robot setups.
• Build tools to inspect robot episodes, annotate failures, and understand why a rollout succeeded or failed.
• Automate dataset quality checks and validation reports for robot data.
• Implement evaluation workflows for testing robot tasks on hardware.
• Track data from physical rollout through training and evaluation.
• Debug issues across robot hardware, sensors, teleoperation, data logging, and learned policies.
• Work closely with robotics and ML engineers to turn failures into better data and better experiments.

Requirements

• Current or recent degree candidate in robotics, mechanical engineering, electrical engineering, computer science, or a related field.
• Strong Python and comfort working with robotics software.
• Hands-on experience with a physical robot, hardware project, lab system, or sensor-based system.
• Experience debugging real systems where software, hardware, and data interact.
• Some experience training, evaluating, or deploying an ML model.
• Evidence of independently building difficult technical projects.

Skills

Bonus points if you have:

• Experience with robot data collection, teleoperation, ROS2, sensors, cameras, calibration, or hardware-in-the-loop testing.
• Experience building evaluation tools, data viewers, experiment tracking, or dataset validation systems.
• Experience with dexterous hands, manipulation, tactile/force sensing, or bimanual systems.
• A robot, hardware, or data tool that other researchers or engineers have used.

What To Include

• A robotics, hardware, data, or evaluation system you personally built, with code, demo, video, or writeup.
• A real failure you diagnosed: what broke, how you debugged it, and what you changed.

Paid internship | 12 to 16 weeks | On-site | San Francisco, CA