Research Intern: Robust Robotic Manipulation

Honda Research Institute USA; San Jose (CA); Presencial; USD 30.000 - 39.000; Jornada completa

Descripción de la vacante

Honda Research Institute USA in San Jose, CA is seeking a self-motivated Ph.D. student to join the Intelligence Robotics Research Division as a summer intern.

The intern will conduct research on robust, contact-rich dexterous manipulation, with a focus on methods that detect, prevent, and recover from failures of learned policies. San Jose, CA Key Responsibilities

Formación

• Currently enrolled in a Ph.D. program in Computer Science, Robotics, Electrical or Computer Engineering, Machine Learning, or a related field.
• Strong background in robot learning and core robotics concepts (e.g., kinematics, dynamics, control).
• Research experience in at least one relevant area, such as failure detection, recovery methods, or policy fine-tuning.
• Proficiency in Python and experience with robotics physics simulators (e.g., MuJoCo, Isaac Sim).
• Ability to independently formulate research questions, design experiments, and analyze results.
• Strong written and verbal communication skills and the ability to work effectively in a collaborative research environment.

Responsabilidades

• Formulate and develop novel algorithms for robust manipulation, including policy performance characterization, failure reasoning, post-training (e.g., RL fine-tuning), and recovery policy methods.
• Develop methods that utilize multimodal sensing (vision, proprioception, tactile) to prevent and recover from policy failures.
• Implement and validate methods in simulation and on real robot hardware.
• Contribute to HRI's patent portfolio and publish research results when applicable at top-tier conferences and journals in robotics and machine learning.
• Communicate progress through presentations and written reports.

Conocimientos

Python

Robot learning

Kinematics

Dynamics

Control

Educación

Ph.D. student in Computer Science, Robotics, Electrical or Computer Engineering, Machine Learning

Herramientas

MuJoCo

Isaac Sim

Descripción del empleo

Research Intern: Robust Robotic Manipulation - Honda Research Institute USA

Research Intern: Robust Robotic Manipulation

Job Number: P26INT-26

Honda Research Institute USA (HRI-US) is seeking a self-motivated Ph.D. student to join the Intelligence Robotics Research Division as a summer intern. The intern will conduct research on robust, contact-rich dexterous manipulation, with a focus on methods that detect, prevent, and recover from failures of learned policies.

San Jose, CA

Key Responsibilities

• Formulate and develop novel algorithms for robust manipulation, including but not limited to policy performance characterization, failure reasoning, post-training (e.g., RL fine-tuning), and recovery policy methods.
• Develop methods that utilize multimodal sensing (vision, proprioception, tactile) to prevent and recover from policy failures.
• Implement and validate methods in simulation and on real robot hardware.
• Contribute to HRI's patent portfolio and publish research results when applicable at top-tier conferences and journals in robotics and machine learning.
• Communicate progress through presentations and written reports.

Minimum Qualifications

• Currently enrolled in a Ph.D. program in Computer Science, Robotics, Electrical or Computer Engineering, Machine Learning, or a related field.
• Strong background in robot learning and core robotics concepts (e.g., kinematics, dynamics, control).
• Research experience in at least one relevant area, such as failure detection, recovery methods, or policy fine-tuning.
• Proficiency in Python and experience with robotics physics simulators (e.g., MuJoCo, Isaac Sim).
• Ability to independently formulate research questions, design experiments, and analyze results.
• Strong written and verbal communication skills and the ability to work effectively in a collaborative research environment.

Bonus Qualifications

• Experience with reinforcement learning, imitation learning, multimodal learning, and/or state estimation.
• Experience deploying learned policies on real robots, including sim-to-real transfer and real-world data collection.
• Experience with dexterous multi-fingered hands, contact-rich manipulation, and/or tactile sensing.
• Publications at relevant venues such as ICRA, IROS, CoRL, RSS, CVPR, NeurIPS, etc.

Years of Work Experience Required 0

Desired Start Date May 2027

Internship Duration 3 Months

Position Keywords Robotic Manipulation, Robot Learning, Failure Reasoning, Multi-modal Sensing