Creating a Robust and Expandable Framework for Cooperative Aerial Robots

Abstract

This work focuses on the design and development of a microservice-based software framework for multi-drone autonomous systems, that is fault-tolerant, expandable, and easily monitored. The custom framework utilizes the Robot Operating System (ROS) and is applied on onboard embedded devices to command and control multiple unmanned aerial vehicles (UAV) in various scenarios, including autonomous flight missions, inter-process communication, as well as measurement fusion for mapping and localization purposes. Further, the proposed framework is implemented and tested in both hardware-in-the-loop simulations and real-world field tests to demonstrate and evaluate its performance. Specifically, the developed prototype is evaluated in scenarios where a swarm of UAVs is tasked to map a predefined area for search-and-rescue purposes and a scenario incorporating faults where one UAV loses its GPS signal and obtains relative positioning with the help of the rest of the UAVs in the swarm.