What does it take to move from maps to momentum? At the intersection of satellite data, AI, and urgent real-world challenges, a new wave of innovators is answering that question – and transforming how we plan, respond, and act.
At the latest session of the TGI GeoAI Working Group, three award-winning teams from the TGI-AWS Generative AI Challenge shared how they are using geospatial AI to drive impact – from wildfire recovery and deforestation monitoring to linking land use with public health. This was the final session in a four-part series showcasing all ten winning teams selected from over 100 proposals earlier this year.
Moderated by Marge Cole, TGI’s Strategic Initiatives Director, the May 1 meeting featured the next and the final cohort of winners from the Challenge: Darryl Murdock and Maximilian Castelli from Avineon, Andrew Kalukin from Through Sensing, and Mikel Maron from Earth Genome. In her opening remarks, Cole welcomed attendees and highlighted the diversity and advancement of the projects, thanking participants for contributing to the continued momentum of the Challenge.
Each team showed what’s possible when geospatial data meets the generative power of AI as working solutions to real-world problems.
Reports from our previous Geo AI Working Group Meetings:
Avineon: Real-Time Tools for Wildfire Damage Response
Darryl Murdock and Maximilian Castelli from Avineon kicked off the session by sharing their wildfire recovery decision support prototype. Built on the AWS SageMaker platform, the tool fuses post-wildfire burn severity data, 3D building information, and risk analysis to help governments prioritize rebuilding efforts.
The dashboard provides agencies with insight into where buildings were damaged, which areas are safe to access, and where to direct resources first. The goal? Speed up recovery in the aftermath of devastating wildfires by making better decisions, faster.
Murdock emphasized the need for interoperable, scalable tools that can integrate with existing workflows across agencies and jurisdictions. “We want to make this accessible to decision-makers who don’t have time to sort through raw data,” he said. “The power of GeoAI is that it can turn complexity into clarity.”
Through Sensing: Early Warning for Deforestation and Biodiversity Loss
The second presentation came from Andrew Kalukin of Through Sensing, who showcased a prototype designed to detect early signals of forest degradation — a key precursor to deforestation and wildfire outbreaks. The tool blends multispectral and Synthetic Aperture Radar (SAR) satellite imagery with generative AI models to analyze and anticipate changes in forest ecosystems before they become apparent through traditional monitoring.
Kalukin explained how the system leverages datasets like ESA Sentinel and Copernicus Land Cover to train AI models capable of identifying subtle signs of ecological stress. The goal is to give forest managers and policymakers a proactive, AI-driven decision support tool.
“The biggest value here is proactive response,” Kalukin noted. “We are not just looking at what happened. We’re asking what might happen — and why.”
Kalukin is part of a collaborative team led by Dr. Dong Xu of the University of Missouri–Columbia, and including researchers from other universities.
Earth Genome: Tracking Land, Food, and Health with AI
Mikel Maron, Product Lead at Earth Genome, closed out the session with a presentation on Earth Index — a generative AI-powered platform that helps map environmental changes and connect them to downstream impacts on water quality and public health.
Using satellite imagery and LLMs, Earth Genome is able to detect land use changes such as the development of pineapple plantations in Costa Rica, then cross-reference that with disease outbreaks and water data to identify potential correlations. The goal is to give local governments and communities better tools to respond to the complex ripple effects of environmental change.
“The patterns are there, but they’re buried in disconnected data streams,” Maron explained. “What we are doing is making those patterns visible — and actionable.”
Collaboration at Core
This final installment in the TGI-AWS Gen AI for Geospatial Challenge series highlighted just how diverse — and impactful — the applications of GeoAI can be. Each project exemplifies the collaborative, problem-solving spirit of the AWS Challenge and the TGI GeoAI Working Group, where academic institutions, private companies, and public sector teams are co-creating solutions that reach beyond traditional boundaries.
Continue to follow these organizations as these projects evolve – and as new voices and innovations emerge from this growing GenAI ecosystem.
About Taylor Geospatial Institute
TGI is passionate about fueling geospatial science and technology to create the next generation of solutions and policies that the whole world will depend on for sustainability and growth.
The TGI consortium includes Saint Louis University, the Donald Danforth Plant Science Center, Harris-Stowe State University, University of Illinois Urbana-Champaign, Missouri University of Science & Technology, University of Missouri-Columbia, University of Missouri-St. Louis, and Washington University in St. Louis. Collectively, these institutions cover geospatial research from ocean depths to outer space.
For more information, visit taylorgeospatial.org.