Detailed Agenda

This presentation demonstrates how GeoAI and ArcGIS Online/ArcGIS Pro can be used to understand the spatial drivers of unsheltered homelessness in California. Using object detection on Google Street View imagery, tents are identified and counted to establish ground-truth data. These observations are then analyzed with socioeconomic variables and geographic assets such as shelter access, transit proximity, parks, and heat exposure to uncover environmental and structural determinants of homelessness. AI and spatial analytics reveal vulnerability patterns guiding equitable policy.

This study uses GIS and machine learning to examine how socioeconomic, demographic, environmental, and structural (geographic asset) factors shape gun violence across U.S. communities. By integrating spatial predictors such as income, education, population density, and proximity to schools, parks, and liquor outlets, the model identifies geographic patterns that explain where violence occurs. The analysis further estimates gun violence risk levels at specific locations, providing actionable insights for targeted prevention and policy intervention.

This presentation will preview how the Smithsonian National Museum of Natural History (NMNH) is transforming 148M+ specimens into a spatially enabled research and education resource. By taking the geographic approach to collection management, the museum is enabling new insights on biodiversity, conservation, climate change and more by unlocking access to the full breadth of their vast scientific collection. Attendees will learn how a GIS platform can support collaboration, digitization, and data-driven discovery across science and education.