The last decade the MAVs have steadily gained interest in the fields of real-life applications, such as infrastructure inspection [1], underground mine tunnel inspection [2], and bridge inspection [3]. The key objective in all these usecases is the online collection of critical information like images, 3D models, and other sensorial data to create safer conditions for the personnel while reducing the overall inspection time. Fully autonomous performance of the MAVs is one of the key challenges in deploying them in real-world applications, while it should overcome uncertainties in localization, limited on-board computation power, delays in control layers, dynamic/static obstacles, etc. Meanwhile, advances in technologies, such as 5G [4] telecommunications technology, enable the use of cloud and edge computing for MAV applications. In this case, the heavy computational processing for multiple processes, such as mapping, localization, and path planning can be carried on the edge computing side, while retaining a fast bi-directional link with the MAV. However, one of the main challenges in such networked applications is the limited bandwidth, the time delays, and the overall package losses that degrade the overall control performance and could lead the system to instability.