As AI becomes deeply embedded in every layer of our digital infrastructure, from industrial automation and smart cities to autonomous vehicles and healthcare, the underlying computing fabric must evolve to support pervasive, data-driven intelligence. Moreover, as billions of Internet of Things (IoT) devices generate ever larger volumes of data, sending everything to distant cloud data centers creates significant bottlenecks for real-time applications due to high network latency, exposure of sensitive data, and the environmental cost of power-hungry cloud facilities.

The edge-to-cloud continuum addresses these limitations by pushing part of the computation closer to where data is generated – on IoT devices or nearby edge servers. However, deploying complex distributed systems such as streaming analytics, federated and distributed machine learning, and AI-enabled services on heterogeneous, mobile, and fault-prone edge infrastructures introduces a new set of challenges. These range from selecting appropriate hardware among diverse options, purchasing and configuring large-scale testbeds, and managing transient faults to coping with limited observability and incomplete runtime insights.

In this talk, Dr. Symeonides will present a methodology for rapid prototyping, benchmarking, and emulation of edge-to-cloud infrastructures tailored for data-intensive and AI-driven applications. By leveraging low-cost, large-scale emulated testbeds that accurately mirror key edge-to-cloud behaviors, we enable reproducible, scalable, and workload-agnostic evaluation. Furthermore, he will demonstrate how this approach guides the optimization of streaming analytics and federated learning pipelines under realistic network conditions and constrained resources. Finally, Dr. Symeonides will present early results on energy- and carbon-aware systems for AI-enabled cyber-physical infrastructures, incorporating models of renewable energy generation, grid carbon intensity, and edge-device power consumption to support sustainable AI execution at the edge.

Dr. Moysis Symeonides is a postdoctoral researcher in the Laboratory for Internet Computing (LInC) at the University of Cyprus (Computer Science Department). His work focuses on edge computing and Internet-of-Things applications, with emphasis on modeling, emulation, experimentation, and performance evaluation of IoT data intensive workloads on 5G/6G edge infrastructures. He received his Ph.D. in Computer Science from the University of Cyprus in 2022, an M.Sc. in Information Systems from Aristotle University of Thessaloniki (2015), and a B.Sc. in Computer Science and Biomedical Informatics from the University of Thessaly (2012).

Since 2014, he has worked as a researcher in many EU projects (PasSPort FP7 EU project, Unicorn H2020, RAINBOW H2020) and tech-oriented companies. He has secured over €190K in European Commission funding and has received awards at leading IEEE/ACM venues, such as IEEE CloudCom (Best Paper, 2024), IEEE/ACM UCC (Best Paper, 2023), IEEE IoTDI (Best Paper, 2022), IEEE ISCC (Best Student Paper, 2022), and ACM/IEEE SEC (Best Demo, 2020).