Stanford Energy Student Lectures - AI Enhanced Battery Modeling, Community Solar Policy, and AI-Driven Materials Discovery

Join us as grad student/postdoc speakers from various technical areas deliver short and accessible presentations about their innovative clean energy research. Learn more about cutting-edge science and the most recent breakthroughs in areas such as renewables, energy conversion materials and devices, catalysis, and decarbonization from the researchers themselves!

Refreshments will be provided starting at 2:45pm. Share your feedback on the speakers for a chance to win a Coupa gift card!

Speaker Topics:

Xiaoyu Yang - Bridging Physics, Data, and Learning in Lithium Battery Modeling

Abstract:

Battery modeling plays a critical role in the design, control, and safety of next-generation energy storage systems, yet significant challenges remain in achieving both physical fidelity and computational scalability across multiple length scales. This presentation discusses a research framework that integrates physics-based modeling, observation-driven correction, and machine learning to address these challenges. The work spans electrochemical transport within porous electrodes, state estimation under model uncertainty, and scalable surrogate modeling for battery systems operating under thermal gradients. A dual-continuum electrochemical transport formulation is introduced to capture transport-limited behavior beyond conventional electrode assumptions. The presentation also examines how indirect measurements can be combined with imperfect models to improve estimation of internal battery states, particularly core temperature. In addition, a machine-learning-based surrogate framework is presented for efficiently modeling coupled electrochemical–thermal interactions in multilevel battery systems. Together, these studies demonstrate how physics, data, and learning can be combined to develop accurate, efficient, and scalable battery models for applications including fast charging, thermal management, diagnostics, and large-scale battery system optimization.

Speaker bio:

Xiaoyu Yang is a Ph.D. candidate in Energy Science and Engineering at Stanford University. She works with Prof. Daniel Tartakovsky and Prof. Yi Cui on physics-based modeling and data-driven simulation of lithium batteries. Her research focuses on balancing physical fidelity, computational efficiency, and scalability across multiple length scales to support battery system design, operational optimization, and engineering decision-making.

 

Ireri Hernandez - Designing Fair Agreements for Solar Energy: What Communities Want from Large-Scale Solar Development

Abstract:

Large-scale solar energy projects are central to the clean energy transition and to increasing electricity demands. However given their size and proximity to residents, some projects may face delays at the local siting and permitting stage. One proposed solution has been Community Benefit Agreements,  which are meant to address local concerns by specifying what host communities receive in exchange for new development. Little is know about which specific benefit and procedural features are likely to increase local support. This project presents results from a survey experiment on public preferences over Community Benefit Agreements for large-scale solar projects. Respondents in areas with existing or potential for large solar projects evaluated alternative agreement designs that varied in both content and process. Specifically, individuals rated projects based on compensation, local jobs, land-use provisions, public participation, transparency, facilitation, and monitoring. The results suggest that both benefits and process matter to increase public support. On the benefit side, respondents were more supportive of agreements that included compensation, local economic benefits such as supply-chain provisions, and environmental management measures. Additionally, respondents were more supportive of agreements negotiated through open meetings, broad resident participation, and ongoing public oversight. These patterns point to a simple but important lesson for solar deployment: community agreements are not only about what is offered, but also about whether the process feels credible and fair.

Speaker bio:

Ireri Hernandez is a Postdoctoral Researcher at Stanford University's Doerr School of Sustainability, where she is part of the Environmental Social Sciences group. Her research examines how the energy and climate transitions shape socioeconomic outcomes and public preferences. Her recent work focuses on how energy prices and economic security influence sustainability goals and public support for the clean energy transition. She employs quasi-experimental methods, survey experiments, and large geocoded datasets to address these questions empirically.

 

Sreya Vangara - From Battery Data to Discovery: AI Agents for Energy Materials Research

Abstract:

Batteries are central to a clean-energy future, but understanding how they work and why they fail increasingly requires reasoning across overwhelming experimental evidence. A single battery study can produce electrochemical cycling data, spectroscopy maps, microscopy images, materials characterization, lab notes, and relevant scientific literature. The bottleneck is no longer simply collecting data; it is turning fragmented evidence into insight quickly enough to guide the next experiment.

In this talk, I will present an AI agent tool we are developing for battery materials research and testing on real experimental datasets at SLAC. The system helps researchers ask natural-language questions across multimodal data and literature, extract quantitative trends, connect observations to physical mechanisms, and identify useful follow-up experiments. Rather than replacing scientists or automating the laboratory, the goal is to build AI systems that serve as scientific partners: tools that help researchers reason across data types, length scales, and hypotheses.

Using battery experiments as a case study, I will show how agentic AI can make complex evidence more searchable, interpretable, and actionable. The broader vision is to accelerate next-generation energy storage discovery by helping scientists ask sharper questions and make better decisions from the data they already collect.

Speaker bio:

Sreya Vangara is a PhD candidate in Mechanical Engineering at Stanford University, where she works at the intersection of AI for science, battery materials, and energy systems. Her research develops AI agents that integrate multimodal experimental data, structured databases, and scientific literature to support battery materials analysis and experiment planning. She has hands-on experience building and testing battery cells, conducting spectroscopy-based materials characterization, and developing AI/ML systems for scientific discovery, including work connected to SLAC energy materials research. She is also contributing to DOE Genesis/CM2US efforts on AI-accelerated materials discovery and national energy-security challenges, including work on computational infrastructure and AI workflows for critical materials research.

Sreya is a Knight-Hennessy Scholar, NSF Fellow, and Quad Fellow. Before Stanford, she earned a master’s degree in International Relations as a Schwarzman Scholar at Tsinghua University and worked on science and technology policy topics including critical minerals, semiconductor industrial policy, and climate resilience. Her broader interests center on using science, engineering, AI, and policy to accelerate clean energy technologies and strengthen national energy security.