Research Scientist Intern, Robot Learning & Manipulation (Spring, Summer & Fall 2027) | Master’s & PhD Internship

Location: Irvine, CA Department: Engineering Employment Type: Internship Location Type: Onsite

FieldAI is where embodied AI meets real robots, real sensors, and real field deployments. We build risk-aware, reliable, field-ready AI systems that tackle challenging robotics problems and unlock the potential of embodied intelligence. Our approach combines frontier research with rigorous engineering, connecting advances in robot learning with the demands of physical systems operating in unpredictable environments.

About the Internship

We are offering spring, summer, and fall 2027 research internships in the Irvine office for master’s and PhD students advancing robotic manipulation.

You will investigate how robots can perform useful physical tasks reliably across diverse objects, environments, and robot platforms. Working closely with researchers and engineers, you will own a focused research project and help translate promising ideas into capabilities that can be tested on hardware and improved through real deployment experience.

Research directions and methods will be shaped by your expertise and the problems we encounter in the field. The goal is to produce original research with measurable impact on what our robots can accomplish.

What We Offer

Real deployment problems: Work on manipulation challenges grounded in customer use cases and operational needs, with opportunities to evaluate beyond controlled laboratory settings.

Physical robots and hardware: Develop and validate your work on robotic systems, including humanoids, mobile manipulators, and arm-equipped quadrupeds.

Research ownership: Shape a focused project, propose your approach, and carry it through implementation, experimentation, and hardware validation.

Deployment feedback: Use observations and failure cases from real operation to guide research and improve reliability.

Close technical collaboration: Work with researchers and engineers across robot learning, perception, control, hardware, and deployment.

Research contributions: Pursue work suitable for leading conferences and journals, with opportunities to release code, datasets, or other research artifacts.

What You Will Get To Do

Develop manipulation capabilities that address meaningful physical tasks in real environments.

Investigate how skills transfer across objects, tasks, environments, and robot platforms.

Address practical challenges such as uncertainty, contact, changing conditions, and failure recovery.

Design experiments that measure task success, generalization, and reliability under realistic operating conditions.

Integrate and validate research on physical robots, collaborating with engineering teams to address system constraints.

Document findings, present results, and contribute to publications or reusable research tools.

What You Have

Current master’s or PhD enrollment in Robotics, Computer Science, Electrical Engineering, Mechanical Engineering, AI/ML, or a closely related field.

Research experience in robotic manipulation, robot learning, embodied AI, or a related area.

A strong research or open-source track record, demonstrated through substantive publications, research software, or technically rigorous projects.

Clear ownership of original research or technical development, including implementation and evaluation.

Strong foundations in robotics, machine learning, and experimental design.

Proficiency in Python and/or C++.

Ability to work independently, communicate findings clearly, and collaborate across research and engineering disciplines.

The Extras That Set You Apart

First-author publications at venues such as RSS, CoRL, ICRA, IROS, NeurIPS, ICLR, or ICML, or journals such as RA-L, T-RO, or IJRR.

Significant contributions to open-source robotics or machine learning projects with demonstrated adoption.

Hands-on experience taking research from an initial idea to reliable operation on physical robots.

Experience evaluating robotic systems outside controlled laboratory conditions.

Public research code, datasets, or robot demonstrations showing technical depth and measurable results.

Experience with robot integration, sensing, control, or data collection infrastructure.