Geospatially Grounded LLM Agents This project gives large-language-model agents access to spatial context so they can reason about terrain, infrastructure, communities, and evolving wildfire conditions. It moves agentic AI beyond text-only inference by treating location as a first-class reasoning layer.
LLM-Augmented Semantic Digital Twins This research develops semantic digital twins that connect technical documents, system models, and evolving project knowledge for adaptive infrastructure planning.
LLMs as Human-Centered World Models This project explores LLMs as data-driven world models for simulating how communities and infrastructure may experience a disaster before it occurs.
