Research to enable advanced battery data analytics using multi-modal AI agents was recognized as among the six best posters at the AI x Energy Summit. Sandia’s Sel Ly presented the poster titled, “Multi-modal AI-Agents for advanced battery data analytics” at the summit, held in San Diego, CA, USA on March 19-20, 2026. The research was a collaboration with contributors to the QuESt open-source platform, developed by Sandia National Laboratories.
The work enables advanced battery data analytics by integrating multi-modal AI agents that can interpret diverse data sources (PDF, DOCX, TXT, CSV, XLSX, and Parquet), automate complex analysis, and provide actionable insights to accelerate energy storage research, optimization, and deployment. It also enables probabilistic battery degradation estimation supporting multiple chemistries (LFP, NCA, NMC) under real-world distribution and domain shifts, directly improving battery management reliability.
As part of the QuESt platform, the research can improve electricity’s affordability by allowing utilities, planners and operators to optimize energy storage deployment and operation to reduce system costs, lower peak electricity prices, and ultimately decrease electricity bills for consumers. The work is also relevant to energy storage companies, data center operators, grid operators, battery manufacturers, and researchers in energy storage and AI. Intelligent, adaptive tools capable of interpreting complex battery datasets could support next-generation data centers and future energy systems.
The work has also been accepted as a paper scheduled to appear at the upcoming 2026 IEEE Power & Energy Society General Meeting: Sel Ly, Dilip Pandit, Preger Yuliya, Tuan A. Ho, Tu. A. Nguyen, Raymond H. Byrne, “AI Agent for Probabilistic Capacity Degradation Estimation for Varying Lithium-ion Batteries”, The 2026 IEEE Power & Energy Society General Meeting (PESGM), pp. 1-5.
This material is based upon work supported by the U.S. Department of Energy, Office of Electricity (OE), Energy Storage Division.