Aerial AId Closeout Report

This report provides an overview of the Aerial AId activity, which focused on enabling assured computer vision onboard Uncrewed Aerial Systems (UAS) to deliver rapid situational awareness for emergency response operations. Conducted under the Convergent Aeronautics Solutions (CAS) project within NASA’s Aeronautics Research Mission Directorate, Aerial AId emerged from the CAS Discovery process in fiscal year 2024 and was executed through fiscal year 2025. The project aimed to address capability gaps in Drone-as-a-First-Responder (DFR) operations by developing and assuring trustworthy AI/ML perception engines tailored for aerial medical emergency scene assessment. Key accomplishments included the development of a runtime assurance framework for computer vision, the creation of a machine learning dataset focused on human stance detection for medical UAS applications, and the training of prototype object detection algorithms using aerial imagery. Aerial AId advanced the state of trustworthy computer vision technologies for autonomous DFR operations, contributing capabilities that support NASA’s mission to develop aeronautics technologies for societal benefit. This report documents the project’s formation, planning and execution, technical approach and outcomes, stakeholder engagement, and lessons learned—offering recommendations for future development of emergency response UAS technologies.