About the role#
This internship focuses on learning-based locomotion and planning for PhD students. You will work at the intersection of reinforcement learning, locomotion control, and learned planning. The goal is to develop integrated systems that allow robots to navigate complex, unstructured environments. You will collaborate with research scientists and engineers to design experiments, build locomotion systems, and validate your work in simulation and on real hardware.
What you'll do#
- Design, implement, and evaluate reinforcement learning pipelines that integrate locomotion control with learned planning.
- Explore how learned planners adapt locomotion behaviors across varied terrain and dynamic conditions.
- Develop sim-to-real transfer strategies, including domain randomization, system identification, and adaptive methods.
- Build and use GPU-accelerated simulation environments like Isaac Gym, Isaac Lab, or MuJoCo for training and evaluation.
- Test and iterate on policies using real legged robot platforms.
- Translate research concepts into working robotic systems.
- Develop experimental setups and tooling to support data collection and reproducibility.
- Work 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.
- Prototype quickly, run experiments, and analyze results rigorously.
- Debug system-level issues across simulation, software, hardware, and learning.
What you'll need#
- Current PhD student in Robotics, Computer Science, Mechanical Engineering, AI/ML, or a 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 or C++ with experience using robotics or ML tooling.
- Familiarity with physics-based simulators like Isaac Gym, Isaac Lab, MuJoCo, or PyBullet.
- Experience designing experiments and evaluating results on robotic systems.
- Curiosity, initiative, and an interest in building autonomous systems for the real world.
Location & details#
- Location: Irvine, California
- Modality: On-site
- Commitment: Full-time internship
About FieldAI
FieldAI develops general robot intelligence for physical environments. The company creates software that allows robots to function in unstructured spaces without GPS or predefined routes. Its technology works with various hardware types across the construction, energy, security, and manufacturing sectors. Founded in 2023, the firm maintains its headquarters in Irvine, California. It currently employs 282 people.
How to get in at FieldAI
Securing an internship at FieldAI requires speed, as early applicants often move through the process before the recruiter's inbox becomes unmanageable. Intern Insider sends an instant alert the moment a role matching your target is published anywhere, so you apply among the first. Applying early ensures your profile is seen before the pile grows. You can also improve your chances by reaching out to the right people. Intern Insider surfaces the recruiters behind the company's roles so you can contact them directly to ask about the position or a referral. This approach often improves response rates compared to submitting an application into a general queue.
