ResQDrone: Autonomous Indoor Quadcopter for Multi-Layered Mapping in Crisis and Disaster Management Scenarios

This paper introduces a fully autonomous quadcopter developed for the safe and efficient exploration of indoor environments in crisis and disaster management scenarios. These environments are typically partially destroyed, entirely unknown, and suffer from restricted visibility. During exploration, multiple mapping layers are generated and transmitted to human operators in real-time, delivering critical insights into the structure. These layers include geometric, semantic, and thermal maps, each contributing distinct information: the geometric layer provides a precise volumetric layout; the semantic layer enriches this layout with contextual understanding of different structural classes; the thermal layer highlights relevant signatures - e.g., humans - that support timely and informed intervention. Successful tests in controlled and simplified real-world environments demonstrate the system’s robustness and its practical value in operational contexts.