News Release

University of Oklahoma to use AI to predict tree failures before storms hit

$1.2M NSF grant funds community-centered framework to reduce infrastructure damage and improve emergency response

Grant and Award Announcement

University of Oklahoma

NORMAN, Okla. – An interdisciplinary team of researchers from across the University of Oklahoma has received a $1.2 million grant from the National Science Foundation to transform how communities anticipate and mitigate risks from treefall during extreme weather events. Led by Aikaterini Kyprioti, an assistant professor from the School of Civil Engineering and Environmental Science, the project will integrate advanced technologies with local knowledge to develop solutions that directly benefit people and their communities.

The project’s innovation lies in its AI-driven, scalable risk assessment framework. By combining image recognition from street-view satellite imagery, fragility models that simulate how different tree species respond to stressors, and social and economic impact analyses of treefall disruptions, TREE-CARE will deliver holistic insights into treefall risks. 

Kyprioti describes trees as a sort of double-edged sword: They provide essential benefits to communities, such as shade and improved air quality, but they are not without risk. “We want trees around, but in severe weather, they can cause damage and problems when it comes to emergency response. We want to create guidance that says which trees might be dangerous in a neighborhood so they can be looked at closely.”

The goal of healthy, safe trees will be supported by five coordinated research thrusts: automated hazard detection, structural and fragility modeling of tree failure, risk and impact analysis on infrastructure through economic analysis, behavioral science on public risk perception and decision-making, and policy and management guidance for local implementation.

At the heart of TREE-CARE are Oklahoma communities, some of which — including Norman, Yukon and Ponca City — will serve as testbeds for the project’s framework. By engaging emergency managers, first responders, local Indigenous tribes, local and county officials, nonprofit organizations and homeowners, TREE-CARE will build a participatory environment where solutions are designed with, not just for, communities. Annual workshops, focus groups and collaborative mapping sessions will ensure that local knowledge and traditional ecological knowledge is integrated into the technical framework. Educating the broader community, the project is set to engage children (K-12) and adults through a variety of activities involving, among others, the Science Museum of Oklahoma and the Sierra Club Oklahoma.

Kyprioti said the NSF-funded project was made possible by OU-supported work. A Junior Faculty Fellowship and seed funding from the OU Institute for Community and Society Transformation (ICAST) allowed the team to demonstrate proof-of-concept, engage with initial stakeholders and establish a research agenda.

Long-term, the framework is designed to be adaptable to other regions and hazards, ensuring that the benefits extend far beyond Oklahoma. By weaving together advanced AI, community partnerships and environmental stewardship, TREE-CARE will enable municipalities to evaluate tree conditions, prioritize management strategies, and prepare emergency responses — ultimately reducing downtime, protecting critical services, and delivering lasting scientific and technical benefits.


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.