Architecting the edge-cloud continuum for IoT-based monitoring applications

Monitoring the status and ensuring the integrity of appliances in civil and industrial scenarios in real-time and over time is a top priority worldwide. To achieve this goal, it is necessary to have a strong synergy between multiple tools, disciplines, and approaches, which can be attained through a joint hardware-software co-design of the different components in an Internet of Things (IoT) system. Layered IoT architectures facilitate understanding of the different software components, hardware, sensing, and networking roles of smart IoT applications. These architectures are inherently distributed, spanning from devices installed in the field up to the cloud, passing through intermediary stages at different levels of edge computing infrastructure -- constituting a computational infrastructure known as the edge-cloud continuum. IoT software platforms based on layered architectures are expected to be adaptable to scenarios with varying characteristics, requirements, and constraints from stakeholders and applications. However, they still face challenges, such as the lack of interoperability and managing data across the edge-cloud continuum. A fine balance exists between providing data with minimal delay and satisfying data freshness constraints. The lack of generality also hampers using the same architecture in multiple deployment scenarios. This thesis proposes, implements, and evaluates a multi-layer IoT architecture that is infrastructure-agnostic and designed to meet the challenges of interconnecting system components. The architecture seamlessly interfaces devices, applications, and subsystems through abstractions. It efficiently manages data across the edge-cloud continuum, enabling timely access to data by utilizing customizable proactive edge-caching techniques. Finally, we demonstrate the architecture's versatility by deploying it in different structural health monitoring (SHM) settings with varying requirements, sensing units, and infrastructure configurations.