Job Specifications
About the Internship
Field AI is building the future of autonomy—from rugged terrain to real-world deployment. We’re on a mission to develop intelligent, adaptable robotic systems that operate beyond simulation and thrive in unpredictable environments.
We are offering a Summer 2026 internship focused on learning-based locomotion and planning for PhD students interested in advancing autonomous legged robot capabilities. As a research intern, you will work at the intersection of reinforcement learning, locomotion control, and learned planning, developing integrated systems that enable robots to move and navigate intelligently through complex, unstructured environments.
You will collaborate closely with Field AI research scientists and engineers to design experiments, develop locomotion and planning systems, and validate ideas in simulation and on real hardware. This internship emphasizes building tightly integrated learning-based systems that connect low-level locomotion with high-level planning, translating research into practical, deployable capabilities for real-world robotics.
What You’ll Get To Do
Advance RL-Based Locomotion and Learned Planning Research
Design, implement, and evaluate reinforcement learning pipelines that tightly integrate locomotion control with learning-based planning
Explore how learned planners can inform and adapt locomotion behaviors across varied terrain and dynamic conditions
Contribute to research projects from early-stage ideas through simulation experiments and on-robot validation
Bridge Locomotion and Planning Across the Sim-to-Real Gap
Develop and refine sim-to-real transfer strategies, including domain randomization, system identification, and adaptive methods, for integrated locomotion-planning systems
Build and leverage GPU-accelerated simulation environments (Isaac Gym, Isaac Lab, MuJoCo) for scalable training and evaluation
Test and iterate on policies using real legged robot platforms in unstructured environments
Build Systems That Connect Research to Deployment
Translate research concepts into working robotic systems tested on real hardware
Develop experimental setups and tooling to support data collection, evaluation, and reproducibility
Help ensure locomotion and planning systems are robust, field-relevant, and ready for iterative improvement
Collaborate Across the Full Robotics Stack
Work closely with systems engineers, perception experts, and embedded teams to close the loop between learning and execution
Incorporate real-world telemetry and field data to refine models and improve generalization
Engage with researchers and engineers across the team to align experiments with broader autonomy goals
Rapidly Iterate and Learn
Prototype quickly, run experiments in simulation and on hardware, and analyze results rigorously
Balance exploratory research with concrete deliverables over the course of the internship
Debug system-level issues spanning simulation, software, hardware, and learning
What You Have
Current PhD student in Robotics, Computer Science, Mechanical Engineering, AI/ML, or a closely related field
Research experience in reinforcement learning for continuous control, locomotion, or learning-based planning
Strong foundation in contact dynamics, control theory, and kinematics
Proficiency in Python and/or C++, with experience using robotics or ML tooling
Familiarity with physics-based simulators such as Isaac Gym, Isaac Lab, MuJoCo, or PyBullet
Experience designing experiments and evaluating results on robotic systems (simulation or hardware)
Curiosity, initiative, and a strong interest in building autonomous systems that operate in the real world
The Extras That Set You Apart
Hands-on experience with legged robot platforms (quadrupeds, wheeled-quadrupeds, bipedal systems, or exoskeletons)
Experience with sim-to-real transfer for locomotion or planning policies
Background in learning-based planning, motion planning, or terrain-adaptive control
Familiarity with ROS or ROS2
Publications, preprints, or open-source contributions in locomotion, RL, planning, or control
Experience deploying neural network controllers on resource-constrained or real-time robotic platforms
Interest in bridging cutting-edge research with practical, field-ready robotic systems
Field AI Onsite Work Philosophy
At Field AI, we believe the most effective way to collaborate and solve complex challenges is by working together in person. This is a fully onsite role, and candidates will be expected to work from our Irvine, CA office. In-person engagement is essential to our success, and we offer flexible working hours to support focus and work-life balance.
Why Join Field AI?
We are solving one of the world’s most complex challenges: deploying robots in unstructured, previously unknown environments. Our Field Foundational Models™ set a new standard in perception, planning, localization, and manipulation, ensuring our approach is explainable and safe for deployment.
You wil