Application of artificial intelligence models to solve the problem of losing control over a drone

Abstract

The object of the study is the process of managing a swarm of drones with a possible break in the communication channel with the control hub, which will reduce the risks of losing swarm members in case of loss of communication with the hub, due to the construction of a duplicate control system. The aim of the work is to avoid the drone hanging in an uncertain state when communication with the control hub is lost due to the construction of a redundant control system module based on the IoT Edge Module, which uses a simple pre-trained image classifier to search for a gesture and launches a landing scenario in case of loss of communication with IoT hub. The proposed approach allows the module to be deployed on each member of the swarm, imposing interoperability requirements, and the deployment on each drone allows the autonomous landing scenario operations to be performed independently, in isolation. The use of the IoT Edge deployment model allows building independent modules for state monitoring and control, which makes it possible to build more complex logic of the swarm control system, and the use of a pre-trained recognition model does not require deep knowledge of neural networks. Ref. 12, fig. 6

Keywords

• drone control
• azure
• azure iot hub edge