Multi-Dimensional Internet of Drones Framework for Smart City Applications

Abstract

As urban populations grow, smart cities face increasing demands for efficient management of transportation, public safety, and resource allocation. The use of drone technology can lower delivery costs, especially in rural or hard-to-reach areas where traditional delivery methods are expensive. Drones can provide real-time traffic data, help to monitor traffic events like congestion and enforce traffic rules. In rescue operations drones can be quickly deployed to search for missing persons in large or difficult-to-access areas, such as forests, mountains, or disaster zones. Also, they can be used to provide aerial views of disaster-affected areas and help to assess damage and plan rescue operations. However, realizing their full potential requires addressing technical collaboration challenges. The collaboration between robots and drones represents transformative opportunity. However, this potential comes with a high degree of complexity. The success of collaborative drones relies on addressing critical challenges in autonomy, coordination, information management, regulations, etc. Efficient joint planning of tasks among drones requires advanced algorithms to ensure optimal performance and avoid conflicts. At the same time, drones need a smart system to be able to decide when and how to collaborate and at the same time consider its profitability from the collaboration. In addition, drones’ collaboration opens the opportunities to have real-time situational awareness which can also take into consideration the uncertainty and the trustworthiness of shared information. Equally important drones need to collaborate with existing RWTOs (Real Word Transportation Objects) and comply with norms and regulations related to its operation and collaboration. This doctoral thesis introduces a collaboration framework for drones enabling them to autonomously balance individual and collective objectives while coordinating action plans within a peer network. The framework address challenges related to group awareness via the management of group beliefs. Furthermore, norms between drones and Real Word Transportation Objects (RWTOs) are investigated to enhance drone applications’ flexibility and compliance with regulations. This research proposes a five-stage methodology to develop a collaborative drone architecture, enabling drones to switch between selfish and collaborative behaviors while addressing uncertainty, dynamic regulations, and mission efficiency. The first stage defines the global collaboration architecture and implements a participatory planning approach to demonstrate its effectiveness. The second stage introduces ASBAF (Assessment, Setup, Bidding, Agreement, Feedback), a multi-stage collaboration approach using bio-inspired algorithms and cost/benefit analysis for optimized task allocation. The third stage focuses on group awareness under uncertainty, proposing a belief management approach with belief fusion operators selected via Hierarchical Analysis Process (HAP). The fourth stage enhances flexibility through a runtime norm regulation framework using Multi-Agent Systems (MAS), where Cloud of Norms dynamically adjusts access rules based on environmental factors. The final stage validates the framework via a search and rescue case study, developing a simulation tool with configurable parameters (e.g., drone count, charging stations) and a linear model to assess surveillance effectiveness. The methodology integrates AI, multi-agent systems, and optimization techniques to improve drone collaboration in dynamic environments. This doctoral thesis is composed of six chapters. In Chapter 1, a general introduction is provided, where the research statements, objectives, questions, and methodology are presented. In Chapter 2, the proposed architecture adopted in the collaborative framework for drones is introduced, and its advantages are demonstrated through an implementation example in a transport scenario. Chapter 3 is dedicated to the joint planning of operational tasks within a group of drones. In Chapter 4, the management of group awareness and its impact on collaboration are addressed. Chapter 5 focuses on norm management to enhance the flexibility of drone applications in smart cities, and in Chapter 6, a collaborative drone surveillance system is implemented for search and rescue operations to improve the operational effectiveness of surveillance tasks through better resource management.