Design and Analysis of Cloud Edge Architectures for the Development and Deployment of Distributed AI Application in Health Domain

The work exhibited in this dissertation focuses on Cloud Continuum (CC) to analyze and define different cloud and cloud-edge architectures on which practical applications can be deployed in various areas such as healthcare, the military, cultural heritage preservation, and energy communi- ties. The organization of the entire work consists of a series of case studies divided into three main areas: expert systems and semantics, machine and federated learning, and augmented and virtual reality. Each case study includes the design, implementation, and testing of reference architectures specific to each area. Projects covered include the development of a story composition tool and an expert system for COVID-19 diagnosis based on probabilistic and semantic rules. Other case studies cover the application of deep learning algorithms for analyzing events related to energy communities, implementing a federated supply chain in the military, and air quality monitoring using machine learning and federated learning. Pre-trained models for object recognition in aug- mented and virtual reality contexts are also explored, with practical applications tested at cultural institutions and technology fairs. In addition, the appendix includes a selection of published works that, while not directly central to the thesis, have contributed to shaping and supporting its devel- opment. This research is essential to developing reference architectures for the Cloud Continuum, with significant implications for the evolution of cloud-edge technologies in real-world contexts.