A Real-Time Autonomous Exploration Framework for Indoor 3D Environments Employing Multiple Unmanned Aerial Vehicles

In search and rescue (SAR) operations, rapid and comprehensive exploration of unknown indoor environments is critical for locating survivors and assessing structural integrity. This paper presents a novel multi-unmanned aerial vehicle (UAV) framework for autonomous exploration in GPS-deprived indoor environments, leveraging advanced sensing technologies and algorithmic strategies. The proposed methodology integrates LiDAR and 3D simultaneous localization and mapping (SLAM) for real-time environment reconstruction, coupled with a weighted frontier-based exploration strategy and Dijkstra's algorithm for collision-free path planning. This combination enables UAVs to prioritize unexplored regions systematically while minimizing redundant coverage. The system's efficacy was validated through high-fidelity simulations in RViz and Gazebo, replicating multi-floor damaged buildings. Performance metrics, including total travel distance and percentage of unvisited areas, demonstrate the framework's ability to achieve nearcomplete 3D coverage (exceeding 90 % in tested scenarios) while significantly reducing exploration time compared to manual methods. These results highlight the framework's potential to enhance the safety and efficiency of SAR missions by reducing human exposure to hazardous environments and accelerating critical decision-making.