A High-Fidelity Risk Assessment for BVLOS UAV Rescue Missions in Urban Lakeside Areas

Abstract This study introduces a precise risk assessment approach for UAV-based emergency response at lakesides, utilising 5G connectivity and adhering to SORA 2.5 standards. The system facilitates BVLOS missions via real-time surveillance, AI-powered video analysis, and automated flight planning informed by fluctuating demographic and environmental data. UAVs are centrally administered using a Remote Operation Control Centre (ROCC), facilitating prompt intervention in both temperate and frigid weather situations. Field testing confirms its capacity for decreased response times and enhanced coordination within EU regulatory frameworks. As SORA 2.5 restricts BVLOS operations in populated areas, it does not account for dynamic environmental factors or the presence of controlled safety zones within these areas. This results in overly conservative restrictions that can hinder legitimate, safe UAV operations. Also, current SORA implementations largely rely on qualitative or class-based population risk models. To address this gap, the proposed methodology applies a quantitative, high-fidelity risk assessment using geospatial and real-time data layers like mobile network coverage, detailed population distribution, building structures, and weather forecasts. This enhances the precision of ground risk calculations, enables risk-informed decision-making, and supports regulatory compliance with improved operational flexibility. The proposed methodology aligns with the evaluation, demonstration, and submission phases proposed in recent updates, including tools like ZHAW’s SORA Tool v3.0.0.