Robotics Research, Embodied AI Intern (Spring & Summer 2027)
Confirmed on the employer's own hiring board on Oct 18, 2026. First seen by Alion on Oct 13, 2026. FieldAI scores A on the Alion truth index.
Why Join FieldAI?
FieldAI is tackling one of robotics’ hardest problems: deploying robots in unstructured, previously unknown environments. We are solving one of robotics' hardest challenges: reliable deployment outside the lab. Our Field Foundational Models™ raise the bar for perception, planning, localization, and manipulation, with an emphasis on explainability and safety for real-world use.
You will work alongside a world-class team that values creativity, resilience, and bold thinking. We bring a decade-long track record of real-world deployments, strong performance in DARPA challenges, and experience from organizations such as DeepMind, NASA JPL, Boston Dynamics, NVIDIA, Amazon, Tesla Autopilot, Cruise, Zoox, Toyota Research Institute, and SpaceX.
What You'll Do
Robot Behavior & Learning: Research methods for learning, adapting, and generating robot behaviors from demonstrations, interaction, experience, and real-world data, enabling robots to perform long-horizon tasks in dynamic and unstructured environments.
Active Perception & Embodied Reasoning: Develop methods that enable robots to actively reason about what to inspect, where to observe, and what actions to take next based on task objectives, observations, uncertainty, and environmental context.
Human-Robot Interaction: Develop approaches for intuitive human-robot collaboration, including natural task specification, interactive robot behaviors, human-in-the-loop reasoning, and adaptation to user intent and feedback.
Embodied AI Models: Investigate and adapt foundation models, including vision-language models, vision-language-action models, and world models, for reasoning, planning, interaction, and decision-making on physical robots.
Real-Robot Experimentation: Implement and evaluate research algorithms on robotic platforms operating in realistic, unstructured environments, and use deployment data to iteratively improve the underlying methods.
Research & Publication: Conduct cutting-edge research in embodied AI and robotics, design rigorous experiments and evaluations, and contribute to publications at leading robotics and AI conferences and journals.
Depending on your research background and interests, you may work on one or more of the following areas:
What You Bring
Currently pursuing or recently completed a PhD in Robotics, Computer Science, Aerospace, Electrical Engineering, Mechanical Engineering, or a related field.
Strong research foundation in one or more areas such as robot learning, planning and decision-making, embodied AI, perception, active perception, or human-robot interaction.
Familiarity with state-of-the-art research in embodied AI and robotics.
Strong programming skills in Python and/or C++.
Experience with modern machine learning frameworks.
Experience developing and evaluating algorithms for robotic or embodied systems.
Strong analytical and problem-solving skills, with the ability to formulate research questions and design rigorous experiments.
Strong written and verbal communication skills and the ability to collaborate across research and engineering teams.
What Will Set You Apart
Hands-on robotics experience, including testing and deploying autonomous systems on physical robots. Experience with ROS or ROS2 and real robotic systems.
Publications at leading AI and robotics venues such as CoRL, ICRA, IROS, RSS, NeurIPS, CVPR, or similar.
Research experience in robot behavior and learning, such as imitation learning, reinforcement learning, learning from demonstrations, or learning from interaction and real-world experience.
Research experience in active perception or active inspection, including task-driven perception, information-gathering behaviors, viewpoint selection, or reasoning under uncertainty.
Experience in human-robot interaction, language-conditioned robotics, shared autonomy, or interactive robot learningExperience fine-tuning or deploying large multimodal models, particularly vision-language, vision-language-action, open-vocabulary, or related foundation models.