Job Responsibilities
As a core member of the team, you will be deeply involved in full-stack R&D spanning from algorithmic models to robot hardware, driving the deployment of embodied intelligence systems from laboratory settings to real-world applications:
1. Embodied VLA Model Development: Participate in the architectural design and optimization of Vision-Language-Action (VLA) large models. Lead the fusion and alignment of multimodal information (visual, linguistic, tactile, etc.) to endow robots with the ability to comprehend complex instructions and execute tasks in the physical world.
2. End-to-End Autonomous Operation Systems: Design and implement end-to-end robotic autonomous operation systems. Establish a closed-loop pipeline integrating perception, decision-making, and control, creating efficient architectures that map raw sensor inputs directly to control outputs.
3. Scenario Generalization & Algorithm Innovation: Conduct research on scenario generalization and transfer learning algorithms for unstructured environments. Enhance the robot's capability to perform dexterous tasks—such as grasping, assembly, and manipulation—in previously unseen scenarios.
4. Robot Hardware & Motion Control: Engage in innovative mechanical design of robot embodiments, optimizing configurations based on task requirements. Simultaneously, develop and debug high-precision motion control algorithms, including force control, impedance control, and whole-body coordination.
5. Software-Hardware Co-design & Architecture: Lead the overall planning and selection of robot hardware architecture. Ensure deep synergy between computing platforms, sensors, actuators, and algorithms. Resolve engineering challenges related to embedded deployment, real-time performance, and power consumption.
