Internship
Full Job Description
Role Overview
Join MERL to develop cutting-edge machine learning algorithms empowering robots to learn novel skills from real-world interaction data. You will design, implement, and test advanced AI systems on physical robot hardware, contributing directly to patents and scientific publications.
Key Responsibilities
- Craft ML algorithms enabling autonomous skill acquisition via teleoperation or self-exploration in dynamic environments.
- Pioneer robust methods for handling task variability, developing primitive skills, scene descriptors, and chaining complex manipulation tasks.
- Collaborate with world-class researchers to verify novel theories through rigorous experimentation on industrial robot arms.
About You
We seek a PhD candidate in robotics, CS, engineering, or applied mathematics. Essential qualifications include:
- Mastery of Python and Git version control.
- Deep expertise in reinforcement learning, imitation learning (LfD), computer vision, and classical geometric methods.
- A track record of translating scientific hypotheses into production-grade research code with strong academic credentials.
Priorities
We highly value experience with manipulation tasks. Familiarity with Vision-Language-Action (VLA) models like Codex is a significant advantage, as will the ability to navigate U.S. export control regulations and work authorization requirements.
Company
Mitsubishi Electric Research Laboratories
Mitsubishi Electric Research Laboratories (MERL) serves as the premier North American research and development arm of Mitsubishi Electric Corporation.
Practice this interview
A voice mock interview for Internship - Robot Learning at Mitsubishi Electric Research Laboratories, with questions from this job description and your resume, and a scored report afterwards.
Set up interview