Energy Chain Network Research Center
Responsibilities:
1. Conduct embodied-AI algorithm research for the energy and power sector, including but not limited to multimodal perception, motion planning, autonomous decision-making and closed-loop control, with deployment and validation on humanoid robots, quadruped robots or robotic arms.
2. Build an integrated framework of an “AI brain” and digital twins for the energy/power sector, combining LLM/VFM with power-system physical models and real-time operating data, and drive the validation and realization of innovative technologies in real problems.
3. Lead or participate in building embodied-AI datasets and simulation environments; for typical scenarios such as power plants and transmission/distribution, build high-fidelity simulation-training platforms and design reinforcement-learning/imitation-learning pipelines for sim-to-real transfer.
4. Track the international frontier of AI/embodied intelligence; write technical reports, patent applications and project proposals; participate in national/provincial-ministerial energy S&T projects and translate research outcomes into new-type power systems.
Requirements / Qualifications:
1. PhD in Computer Science, Artificial Intelligence, Electronic Engineering, Software Engineering or a related field.
2. Hands-on development experience in reinforcement learning, imitation learning or LLM fine-tuning; successful cases of simulation-environment construction and sim-to-real transfer are preferred.
3. Proficient in Python/C++ and in PyTorch/TensorFlow and ROS/ROS2.
4. CET-6 or an equivalent level of English, able to read English standards and literature fluently; proficient in office software and able to produce professional documents.
5. A fast learner with a strong sense of responsibility; able to collaborate with members from hardware, electrical and communication backgrounds; basic understanding of energy/power business scenarios, sound engineering implementation skills, and an end-to-end mindset from algorithm prototype to engineering deployment.
