Deploying Rapid Damage Assessments from sUAS Imagery for Disaster Response

2025-11-05

AI Summary

In federal disaster response, the volume of sUAS imagery data is immense (47–369 GB per day), manual damage assessment is slow, and existing AI models remain confined to academic validation without operational deployment. Method: This study develops and empirically validates the first end-to-end AI-driven building damage assessment system for hurricane response. It leverages the largest publicly available labeled dataset to date (21,716 sUAS aerial images), integrates computer vision with lightweight machine learning models, and achieves field operational integration with standardized response workflows. Contribution/Results: The system marks the first real-world deployment of AI in FEMA-scale hurricane operations (Hurricanes Debby and Helene), completing damage classification for 415 buildings in 18 minutes—demonstrating a decisive breakthrough in transitioning AI from research to practice. It establishes a reusable technical paradigm and operational standard for AI-augmented disaster response.

Technology Category

• Computer Vision: Remote Sensing / Geospatial AI
• Humans and AI: Planning and Decision Support for Human-Machine Teams
• Intelligent Robots: State Estimation

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

This paper presents the first AI/ML system for automating building damage assessment in uncrewed aerial systems (sUAS) imagery to be deployed operationally during federally declared disasters (Hurricanes Debby and Helene). In response to major disasters, sUAS teams are dispatched to collect imagery of the affected areas to assess damage; however, at recent disasters, teams collectively delivered between 47GB and 369GB of imagery per day, representing more imagery than can reasonably be transmitted or interpreted by subject matter experts in the disaster scene, thus delaying response efforts. To alleviate this data avalanche encountered in practice, computer vision and machine learning techniques are necessary. While prior work has been deployed to automatically assess damage in satellite imagery, there is no current state of practice for sUAS-based damage assessment systems, as all known work has been confined to academic settings. This work establishes the state of practice via the development and deployment of models for building damage assessment with sUAS imagery. The model development involved training on the largest known dataset of post-disaster sUAS aerial imagery, containing 21,716 building damage labels, and the operational training of 91 disaster practitioners. The best performing model was deployed during the responses to Hurricanes Debby and Helene, where it assessed a combined 415 buildings in approximately 18 minutes. This work contributes documentation of the actual use of AI/ML for damage assessment during a disaster and lessons learned to the benefit of the AI/ML research and user communities.

Problem

• Automating building damage assessment from sUAS imagery during disasters
• Addressing overwhelming volume of aerial imagery that delays disaster response
• Establishing operational AI system for rapid damage assessment in emergencies

Innovation

• AI/ML system for building damage assessment
• Training on largest sUAS disaster imagery dataset
• Deployed during hurricanes for rapid evaluation