Текст заказа
ABOUT THE TEAM AND THE ROLE:
At Mind Robotics, we’re building generalized physical AI—robotic systems capable of dexterous, adaptive, and reasoning-intensive work in real-world industrial environments. Delivering this requires high-quality human demonstration data, and a teleoperation system that lets operators control dexterous robots precisely, quickly, and comfortably.
We’re looking for a Tech Lead, Teleoperation to build the teleoperation stack that powers our data collection—from operator input devices and retargeting to operator-facing tools.
RESPONSIBILITIES:
- Design, develop, and maintain the teleoperation stack that enables reliable, low-latency control of robotic platforms.
- Set the technical direction for teleoperation and lead design and code review across the stack.
- Implement and optimize retargeting and inverse kinematics that map operator motion to robot arms, hands, and grippers.
- Integrate operator input devices like VR headsets, exoskeletons, and motion-capture gloves, or build custom solutions as needed.
- Ensure that teleoperation meets strict latency, tracking accuracy, and safety requirements.
- Record synchronized, high-quality demonstration data for model training.
- Use simulation to validate teleoperation changes before they reach hardware.
- Contribute to operator-facing tools (e.g., UI for calibration, task selection, episode recording, and review).
- Extend the stack from manipulation to whole-body control, and over time to locomotion and navigation.
REQUIREMENTS:
- Strong experience building teleoperation systems for robotic manipulators or humanoids.
- Experience providing technical leadership or mentorship to other engineers.
- Strong understanding of kinematics, retargeting, and robot control.
- Experience working with real-time constraints and performance-critical systems.
- Strong proficiency in Python or C++ programming.
- Familiarity with human motion tracking hardware (e.g., VR/XR headsets, exoskeletons, motion-capture gloves).
- Familiarity with robotics middleware frameworks (e.g., DDS, Zenoh, ROS2).
NICE TO HAVE:
- Experience with dexterous hands, tactile sensing, or haptic feedback.
- Experience with whole-body control, locomotion, or navigation for humanoid robots.
- Experience collecting demonstration data for imitation learning.