MAI Successfully Tested an Unique Unmanned System to Search for Missing Persons
The size of drone is small, which allows to effectively maneuver between trees. Each drone in the swarm will have on board special equipment for autonomous navigation, as well as cameras for shooting and recording video. The captured video attached to the coordinates is processed by the neural network, which determines the presence of people in the frame and transfers the survey to the operator for visual confirmation.
The system has already passed the first field tests, which made it possible to determine the possibility of flying under the tree canopy, depending on the density of the forest. For this purpose, the developers from MAI used racing drones guided through a virtual reality helmet. During the tests, the members of MAI team themselves acted as lost people. Based on the results of the video processing by the neural network, tagged fragments were obtained indicating the probability of the presence of people in them.
The development was implemented by the MAI team within the framework of the Odyssey competition from the Charitable Foundation “Sistema”. The specialists of MAI Drone School participate in the GoFPV team, and with the implementation of the neural network, they are assisted by specialists from the V.A. Trapeznikov Institute of Control Sciences of Russian Academy of Sciences.