A Survey of AI-based Models for UAVs’ Intelligent Control for Deconfliction

Abstract

The growing potential for Unmanned Aerial vehicles (UAVs) is creating new business opportunities. Drones, U-Space, UTM, and their application to Air Mobility are fostering new applications that evolve quicker than the regulatory frame-work. For instance, the number of domains of applications is increasing, ranging from infrastructure inspection to parcel deliveries in urban settings. Thus, the number of drones flying simultaneously in the same geographical area is expected to grow over the next few years. It will soon pose safety issues as it might become more and more challenging to ensure safe control of drones so that they are separated from each other and with manned flight operations. The deconfliction or separation management problem, a pressing issue, has been tackled in several research projects. However, there is still an urgent need for a better approach to automate deconfliction at the strategic and tactical levels. Our SESAR-funded project (AI4HyDrop) aims to explore the use of machine learning to develop an intelligent control system to resolve drone deconfliction. To this purpose, we have conducted an extensive literature review, as outlined in this paper. This research was needed to understand better how AI has been explored in the field of UAV’s deconfliction and thus will pave the way for basic concepts that contribute to the requirement elicitation for an AI model-based UAV’s deconfliction