AeroRescue-AI: A Reinforcement-Driven and Risk-Adaptive UAV Framework for Intelligent Search and Rescue Operations
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
As disaster frequency and complexity escalate globally, the need for intelligent, autonomous, and well-coordinated search-and-rescue (SAR) systems has become increasingly critical. This research introduces AeroRescue-AI, a conceptual framework that leverages unmanned aerial vehicles (UAVs) integrated with reinforcement learning (RL) and data-driven risk assessment to improve the safety, adaptability, and efficiency of SAR missions. Unlike conventional UAV deployments that rely on preprogrammed routes or static response strategies, AeroRescue-AI envisions a dynamic, self-learning system capable of perceiving environmental conditions, assessing situational risks, and autonomously adjusting flight and rescue strategies in real time. The proposed architecture integrates three core components: (i) context-aware reinforcement learning agents that support autonomous decision-making and path optimization, (ii) a risk-aware prioritization model that computes a Rescue Urgency Score (RUS) inspired by data-driven risk prediction techniques, and (iii) a multi-UAV coordination network based on Flying Ad Hoc Networks (FANETs) for distributed sensing and collaborative response. By bridging concepts from intelligent traffic recovery, UAV-enabled public safety operations, and predictive analytics, AeroRescue-AI seeks to provide a coherent blueprint for adaptive, trustworthy, and explainable AI-driven aerial rescue systems capable of operating in complex and uncertain disaster environments.
Fields of Study & Subjects
UAV Applications and Optimization
Energy and Environmental Systems
Robotic Path Planning Algorithms
Publication details
- Journal / source
- International Conference on Consumer Electronics
- Pages
- 1–6
- Publication date
- 3 February 2026
- Language
- EN