Keynote Speakers

RTNS 2026 is pleased to welcome the following keynote speakers, bringing perspectives from both academia and industry on predictability, real-time computing, and the deployment of intelligent systems.

Alessandro Papadopoulos

Professor of Computer Engineering, Technical University of Denmark (DTU) / Professor of Electrical and Computer Engineering, Mälardalen University (MDU), Sweden

Predictability in the Edge-to-Cloud Continuum: When Time Matters

Abstract:

Modern cyber-physical systems increasingly rely on distributed intelligence deployed across devices, edge nodes, fog infrastructure, and the cloud. This computing continuum offers unprecedented flexibility and scalability, but it also challenges one of the central assumptions of real-time systems: that the computational and communication substrate can be characterized with sufficiently stable timing properties. In applications such as autonomous vehicles, industrial automation, and cloud-based control, latency, jitter, resource contention, and missed updates do not merely degrade software performance; they directly affect the behavior of the physical system.

This presentation discusses predictability as a cyber-physical property of edge-to-cloud systems. The talk will argue that predictable cyber-physical intelligence requires moving beyond classical worst-case reasoning alone, toward systems that can understand, expose, and manage timing uncertainty across the computing continuum.

Short Bio:

Alessandro Papadopoulos is Professor of Computer Engineering at the Technical University of Denmark (DTU) and Professor of Electrical and Computer Engineering at Mälardalen University (MDU), Sweden. He serves as Director of the AI@MDU initiative, Vice Director of the Mälardalen University Automation Research Center (MARC), a member of the International Advisory Board of the LUMSA International Research Center for Artificial Intelligence Management, and a member of the IEEE Technical Committee on Real-Time Systems (TCRTS) Executive Committee.

He received his Ph.D. from Politecnico di Milano in 2013 under the supervision of Professor Alberto Leva. In 2014, he joined Lund University as a postdoctoral researcher, working with Karl-Erik Årzén and Martina Maggio on resource allocation for cloud infrastructures and real-time systems. He joined MDU as an assistant professor in 2016 and was promoted to associate professor in 2018 and full professor in 2022.

He has served as Program Chair of MED 2022, ECRTS 2023, ICPE 2025, and SEAMS 2027. He has also held visiting and scientific advisory positions at academic institutions and industrial companies, including the University of Málaga (Spain), the University of Bologna (Italy), ABB (Sweden), and Zero Point Technologies (Sweden).

His research spans the real-time systems, automatic control, software engineering, artificial intelligence, and cyber-physical systems communities. He develops theories, methods, and tools for predictable, adaptive, and resource-aware computing systems operating under uncertainty, with applications in embedded and real-time systems, edge-to-cloud computing, robotics, and industrial automation.

Sergei Chichin

Senior R&D Engineer for Avionics Software, Airbus

From Training to Flight: The Industrial Realities of Embedded AI in Avionics

Abstract

The aerospace industry is increasingly looking to leverage Artificial Intelligence to complement the next generation of flight capabilities. However, taking a Machine Learning (ML) model from an unconstrained training environment and deploying it onto a flying, safety-critical system represents a massive engineering leap. In this keynote, we will explore the industrial realities of Embedded AI from an Airbus perspective, highlighting the friction between modern ML paradigms and the strict, deterministic requirements of aviation.

This presentation will walk through the integration of AI within the Airbus development process, grounded in the emerging ED-324/ARP6983 standard and its "W" development cycle. We will specifically focus on the critical "second V" of this cycle—the deployment phase—where some of the most significant engineering bottlenecks occur.

Attendees will gain insight into the primary challenges of deploying ML inference on hybrid compute platforms, including:

• Embeddability Constraints: Adapting resource-heavy ML models for constrained, high-performance aerospace hardware.
• Predictable Timing: Achieving deterministic execution and worst-case execution time (WCET) guarantees on complex, hybrid architectures.
• Robust Partitioning: Ensuring strict segregation when utilizing modern hardware accelerators.
• Navigating Emerging Standards: Leveraging frameworks and standards such as SONNX for ML pipeline deployment.
• Methods and Tools: Overcoming the complexities of new programming models and adapting to significantly shorter deployment cycles.

By bridging the gap between theoretical ML design and physical hardware deployment, this talk aims to outline open problems and potential future pathways for real-time, safety-critical embedded systems in the avionics industry.

Short Bio

Dr. Sergei Chichin is a Senior R&D Engineer for Avionics Software at Airbus, specializing in high-performance compute platforms for safety-critical real-time systems. With a Ph.D. in Computer Science in the area of combinatorial optimization and more than 10 years of industry experience, his current work focuses on bridging the gap between AI model training and industrial, certifiable avionics deployment. He is a key contributor to related emerging industry standards, including ED-324 and SONNX, with a particular focus on ML inference embeddability, robust hardware partitioning, deterministic execution on hybrid compute platforms, and the Processes, Methods, and Tools (PMT) required for the industrialization of future ML-based avionics applications.