Robots need to make significant progress in manipulation performance and capabilities before they become commonplace in our lives. The Manipulation team is developing robot mechanisms and behaviors that enable smooth, efficient execution of manipulation tasks.
Requirements
- Expertise in machine learning for control, imitation learning, and/or reinforcement learning
- Experience with robot manipulation, motion planning, and perception systems
- Proficiency in Python and C++
- Publications in top-tier robotics or AI venues
- Experience deploying algorithms on real robotic platforms
- Knowledge of dexterous manipulation, task and motion planning, behavior abstractions, skill completion prediction, failure detection and recovery, and/or using machine learning models for long-horizon skill chaining.
Responsibilities
- Researching, developing, and validating new algorithms and models for skill learning, composition, and real-world deployment.
- Collaborate across hardware, software, and research teams to integrate and test these capabilities in physical robotic systems.
- Improving hardware for high-speed, high-strength, and dexterous manipulation, including underactuated manipulators
- Using learning methods to develop robust capabilities for novel environments
- Developing parameterized skills and chaining them into long-horizon behaviors
- Applying robust strategies for failure recovery in manipulation tasks
Other
- PhD or equivalent experience in robotics, computer science, mechanical engineering, or related field
- Master’s degree with several years of industry or applied research experience
- Bachelor’s degree with significant and relevant hands-on industry experience
- 7 Years in Research and Development
- Ability to solve problems rapidly in hands-on, real-world scenarios
