Lawrence Berkeley National Laboratory-posted 2 days ago
Full-time • Entry Level
Hybrid • Berkeley, CA
101-250 employees

Lawrence Berkeley National Laboratory is hiring an Applied Artificial Intelligence & Autonomous Systems Postdoctoral Fellow within the Energy Analysis division. This position resides in the Systems and Energy Technologies Analysis Department and will develop AI-driven autonomous systems for next-generation transportation and grid applications. The role focuses on advanced control, LLM-based agentic frameworks, and robust decision-making to improve the safety and reliability of transportation and grid simulations, with contributions to model development, software maintenance, and real-world deployment. You Will: Develop fine-tuning and/or Retrieval-Augmented Generation (RAG) methods to augment LLMs with dedicated knowledge in transportation and electric grid domains. This involves designing methods to process input data and documents into appropriate structures for efficient training, establishing evaluation metrics to validate model performance, and improving the fine-tuning process with additional reinforcement learning steps. Integrate fine-tuned LLMs into the existing agent-based simulation framework in HEVI-LOAD to enable agentic scenario planning and parameter tuning. Design and develop novel multi-agent reinforcement learning algorithms with safety guarantees for connected autonomous vehicles/trucks, robotic and vehicle-grid integration systems, incorporating robust control methods and uncertainty quantification. Implement and validate safe and/or model-predictive control frameworks for real-time path planning and tracking of autonomous trucking and warehouse management systems. Conduct data analysis on pilot-scale deployments of vehicle-grid integration and develop control systems for smart charging/discharging within the context of both transportation and electric grid systems. Publish research findings in high-impact peer-reviewed conferences and journals, and present findings to prospective collaborators and funders. Assist the PI and other researchers in developing the funding proposals.

  • Develop fine-tuning and/or Retrieval-Augmented Generation (RAG) methods to augment LLMs with dedicated knowledge in transportation and electric grid domains.
  • Integrate fine-tuned LLMs into the existing agent-based simulation framework in HEVI-LOAD to enable agentic scenario planning and parameter tuning.
  • Design and develop novel multi-agent reinforcement learning algorithms with safety guarantees for connected autonomous vehicles/trucks, robotic and vehicle-grid integration systems, incorporating robust control methods and uncertainty quantification.
  • Implement and validate safe and/or model-predictive control frameworks for real-time path planning and tracking of autonomous trucking and warehouse management systems.
  • Conduct data analysis on pilot-scale deployments of vehicle-grid integration and develop control systems for smart charging/discharging within the context of both transportation and electric grid systems.
  • Publish research findings in high-impact peer-reviewed conferences and journals, and present findings to prospective collaborators and funders.
  • Assist the PI and other researchers in developing the funding proposals.
  • Ph.D. in Computer Science, Electrical Engineering, Mechanical Engineering, Robotics, or a closely related field, with a strong research background in AI, reinforcement learning, and multi-agent systems.
  • Demonstrated expertise in agent-based approaches and decision-making algorithms, such as safe and robust control methods, evidenced by publications in top-tier conferences or journals.
  • Strong hands-on programming skills in Python and deep learning frameworks (PyTorch or TensorFlow), with experience in implementing and training complex neural network architectures.
  • Ability to work as an independent researcher with a high level of scientific judgment and initiative, as well as effectively as part of a diverse collaborative team.
  • Strong publication record in reinforcement learning, autonomous systems, robotics, or related fields.
  • Hands-on experience with robotics platforms and sim2real transfer (ROS, Carla, MetaDrive, or similar).
  • Experience in vehicle-grid integration, autonomous vehicles, smart grid systems, or energy storage applications.
  • Experience with vision-language models, large language models, or multi-modal learning systems.
© 2024 Teal Labs, Inc
Privacy PolicyTerms of Service