Urban airspace design for autonomous drone delivery

The paradigm of large-scale adoption of autonomous drone delivery promises to provide commercial and societal benefits. Over the past years, several companies have investigated the use-case of drones to transport small express packages of fast-food meals and time-sensitive medical supplies. The latter has shown to be highly beneficial in many parts of the worldwhere traditional transport infrastructure remains largely non-existent. However, it is assumed that the true value of autonomous delivery drones can only be demonstrated when it is applied to urban environments. For example, the use of a large-scale fleet of autonomous drones to transport packages within the last-mile segment could potentially improve the economics of package delivery, reduce traffic congestion and help decrease the total anthropogenic carbon dioxide emissions in cities. In addition, supplementing the existing last-mile delivery system with this new technology could also help accelerate the European Union’s 2050 vision of de-carbonising the transport sector. Autonomous drone delivery is obviously not a panacea to the above problems. It could, however, offer a path to mitigate such societal problems. Yet, even though there is a compelling case for autonomous drone delivery, it still remains to be deployed in cities. The reasons for this slow adoption include a large number of complex regulatory hurdles that vary between countries and cities. However, the biggest challenge is how to safely harbour large traffic volumes of drones in a constrained urban environment. This thesis frames the scientific problem and outlines twomain past research areas: unconstrained airspace design and road-based design, which served as a rich source of inspiration for this research. In a past study, known as theMetropolis project, it was demonstrated that layering the airspace and allocating flights to different altitude layers with respect to travel directions helped to mitigate the conflict probability in an unconstrained airspace setting. The study revealed two factors that were largely responsible for increasing the level of airspace safety, namely, segmentation of traffic and reduction of the relative speed, by traffic alignment, between cruising traffic at the same altitude. Furthermore, existing road vehicles, especially automated cars, provide an informative comparison with autonomous drones. Both emerging transportation modes are expected to navigate in constrained urban settings and operate in high traffic density scenarios. Of course, there are notable differences, for example, drones will operate in a three-dimensional space and the current performance limits of drones imply that it would not be optimal for drones to come to a sudden halt at intersections unlike cars and thus separating opposing traffic flows at intersections will be a difficult task. Yet, road design and research has evolved alongside road vehicles to include a host of safety measures in effort to make roads and streets safer for all its users. They make use of various conflict prevention measures to structure and organise traffic flows. Current roads and streets have channelisation planes, which help separate opposite flows of traffic using road markings, islands and raised medians to distinguish and support one-way and two-way streets. These forms of structuring have shown to reduce of the risks of conflicts and, to an extent, are able to safely harbour high traffic densities in highly constrained urban environments. The work in this thesis therefore aimed to investigate what design paradigms and methodologies from unconstrained airspace research and road infrastructure design can be translated to a constrained urban airspace for high-density drone traffic operations...