Civilian Drone Services: Concept, Strategies and Experimentation

Submission note: A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy to the School of Engineering and Mathematical Sciences, College of Science, Health and Engineering, La Trobe University, Victoria, Australia.

Recently, drones have attracted increasing attention and are taking mobile computing to a new era. However, the subject of drone service provisioning from the service-oriented computing viewpoint has received relatively little attention and hence, is not well understood. We envision that a variety of drone applications can be delivered as a service. While it is not likely in the near future that everyone will have a drone, nor is it conceivable that there will be as many drones as smartphones today, a range of companies operating drones to deliver particular services might be conceivable in the near future. The broad aim of this thesis is to gain insights into aspects of civilian drone services and to describe and address possible challenges. This work offers three main contributions. First, this thesis elaborates on the concept of civilian drone services. It discusses the case for having community-based and commercial drone services using the Australian city of Melbourne as an example, as opposed to the one-or-more-drones per-person scenario and we explore the factors that may enhance the delivery of drone services. The results suggest that the strategy for selecting which drone to serve a request next, the position and the number of stations, and the number of drones are factors for effective drone services. Second, this thesis proposes and examines programmable crowd-powered drones, involving two key concepts for combining drones and smartphones as a crowdpowered resource cloud. In particular, it focuses on crowdsourcing for drone computations and multi-drone service management using a new scripting language to coordinate the flight paths of multiple drones. The combination of utilising nearby devices (i.e. with additional resources beyond the drone capability) and controlling multiple drones in a more convenient way has the potential to overcome these limitations. Third, this thesis comprehensively studies, via simulation, how different decisionmaking strategies relating to drones impact clients and service providers. In particular, this thesis investigates the trade-offs between maximising provider revenue versus maximising client satisfaction and different combinations of factors that influence drone behaviour. The thesis also investigates a wide range of scenarios to assess the impact of the various combinations of station strategies and a selected drone strategy. The results suggest that a combination of different strategies for task allocation and other external factors achieves radically different results. The research has implications for drone service users of the future.