Tasks:
- * Align data across robot embodiments
- * Coordinate high- and low-frequency control
- * Coordinate mobile base and manipulator control
- * Define embodied data formats and quality standards
- * Design sensor representation and tokenization
- * Develop mixed-data training strategies
- * Develop mobile manipulation architectures
- * Develop multimodal VLA algorithms
- * Enable whole-body control
- * Fuse tactile, force, vision, and language data
- * Improve real-robot generalization
- * Improve real-robot manipulation success
- * Integrate VLA policies with motion control
- * Reduce sim-to-real and demonstration-to-robot domain gaps
- * Train models on demonstration and real-robot interaction data
Perks/Benefits:
Skills/Tech stack required:
[Body Control] [C++] [Co Training] [Cross-embodiment alignment] [Deep learning] [Domain Adaptation] [Force Torque] [Force/torque sensing] [Imitation Learning] [Linux] [Machine Learning] [Mobile manipulation] [Motion Control] [Multimodal Learning] [Python] [PyTorch] [Reinforcement Learning] [Robot dynamics] [Robotics] [Robot Manipulation] [ROS] [ROS 2] [Tactile sensing] [TensorFlow] [Torque sensing] [VLA] [Whole-body control]
Educational requirements:
N/A
Role(s):
[AI Algorithm Engineer] [Algorithm Engineer] [Embodied AI Algorithm Engineer] [Engineer] [Learning Engineer] [Robotics Engineer] [Robot Learning Engineer] [VLA Algorithm Engineer]