07.01.2020, 13:30, room 3945
Abstract:
Every year hundreds of people get buried under avalanches. The survival probability of victims is mainly dependent on the time required to locate the victim, thus a quick triangulation and extrication is desirable. This work addresses the topic of building robust algorithms for search drones that have the potential to save lives. The system builds upon commercially available avalanche beacons which are considered standard equipment when going "off-piste". Equipping a drone with such a sensor, capable of detecting these electromagnetic beacons, an analysis was done to quantify the electromagnetic interference of the host drone's hardware, and the respective interference generated. An evaluation of the recorded data allowed for the localization of avalanche victims. By autonomously maneuvering to the avalanche site and performing a search, the drone was found capable of localizing avalanche victims within a sub-meter granularity. Potentially enabling rescuers to proceed with the extrication faster and hence increases survival probabilities of victims.