Speaker: Valeria Cardellini (University of Rome Tor Vergata)
Title : Adaptive Serverless Computing Across the Cloud-Edge Continuum Abstract: Serverless computing, particularly the Function-as-a-Service (FaaS) model, has transformed application deployment through event-driven execution and automatic scaling. As workloads increasingly span the heterogeneous Cloud-Edge continuum, however, traditional cloud-centric platforms struggle to cope with dynamic node availability, fluctuating network conditions, data locality constraints, and energy budgets.
This keynote explores the paradigm of adaptive serverless computing, where function scheduling, execution offloading, and resource management continuously adapt to changes in infrastructure, workload, and user requirements. We discuss how context-aware orchestration make decentralized runtime decisions that balance performance, cost, and sustainability across heterogeneous cloud and edge resources.
As a case study, we present Serverledge, a decentralized, open-source FaaS platform for distributed continuum environments, showing how adaptive execution enables efficient, resilient, and low-latency serverless performance without centralized control.
Finally, we highlight emerging research challenges and outline a vision for autonomous serverless platforms that leverage self-adaptation and self-organization to orchestrate next-generation workloads across future continuum systems..
Speaker: Carlo Ghezzi (Politecnico di Milano)
Speaker: Marco Dorigo (Université Libre de Bruxelles) Title: Bridging Centralized and Decentralized Control in Robot Swarms through Self-Organizing Hierarchies Abstract: Robot swarms promise scalable and resilient solutions for applications such as environmental monitoring, search and rescue, and logistics, yet their adoption remains limited by poor controllability. Fully self-organizing swarms achieve robustness and scalability through decentralized coordination and simple local interactions, but this comes at the cost of limited observability and difficulty in shaping global behavior. Conversely, centralized approaches offer ease of control but suffer from scalability issues and single points of failure.
In this talk, I present the Self-Organizing Nervous System for Robot Swarms (SoNS), a middleware framework that enables robots to dynamically form adaptive hierarchical structures. SoNS bridges centralized and decentralized paradigms by supporting functionally centralized coordination of sensing, actuation, and decision-making, while preserving the scalability, flexibility, and fault tolerance of self-organizing systems.
As technical systems increasingly observe, decide, adapt, and act with minimal human oversight, the question of trust becomes critical. This panel brings together practitioners to explore what autonomy truly means in real-world production environments, going beyond conceptual models. Through concrete system examples, panelists will examine the gap between theory and practice, discuss where to draw the line between automation and human control, and share evidence-based approaches for verifying correct behavior. They will also candidly address failure modes, uncomfortable trade-offs, and hard-earned lessons that the ACSOS research community should not overlook. Join us for an honest conversation about what makes autonomous systems worthy of our trust—and what keeps practitioners up at night.
Speaker: Ivona Brandic (TU Wien) Title: Bridging Classical and Quantum Computing: Lessons, Challenges, and Opportunities Abstract: Quantum computing promises computational capabilities that extend beyond the representational and computational limits of classical systems. It is widely regarded as a candidate technology for addressing scalability and tractability challenges in domains such as molecular simulation, optimization, and artificial intelligence. However, classical high-performance computing (HPC) systems continue to play a crucial role in the quantum computing ecosystem. Classical computers are essential for tasks such as quantum device calibration, quantum state preparation, error mitigation, and data pre- and post-processing.
In this talk, we critically examine the role of classical computing in conjunction with quantum systems. We revisit early misconceptions about quantum advantage, discuss key lessons learned from the first generation of hybrid quantum–classical systems, and highlight current research directions. Particular attention is given to state-of-the-art hybrid algorithms, including the Quantum Approximate Optimization Algorithm (QAOA) and the Variational Quantum Algorithms (VQA), as well as emerging concepts for tighter integration between classical and quantum infrastructures. We conclude by exploring novel interaction paradigms, such as digital shadows and their practical implementations, which may shape the next generation of hybrid computing systems.
Q&A: 15:35–16:00