Keynote speakers

Prof. Marco Dorigo, Université Libre de Bruxelles

Title: Self-Organizing Nervous System for Robot Swarms

Abstract:

Robot swarms offer the potential for scalable and resilient solutions in domains such as environmental monitoring, search and rescue, and warehouse logistics. However, their practical adoption is often hindered by limited controllability. Numerous studies have shown that fully self-organizing robot swarms can operate without any central coordination mechanism. In these systems, a decentralized structure ensures substantial redundancy, while collective behaviors arise from simple local interactions among individual robots. Although this approach brings recognized benefits—including scalability, robustness through redundancy, and the elimination of single points of failure—it also gives rise to intrinsic limitations, particularly the challenge of monitoring, guiding, and shaping the swarm’s overall behavior. By comparison, centralized approaches are generally simpler to implement and supervise, but they are constrained by poor scalability and vulnerability to single points of failure.

In this talk, I introduce the Self-Organizing Nervous System for Robot Swarms (SoNS), a middleware framework that enables robots to dynamically form temporary and adaptive hierarchical structures. By allowing robots to self-organize into such hierarchies, SoNS effectively bridges the gap between centralized and decentralized control paradigms. Through SoNS, collective sensing, actuation, and decision-making can be coordinated in a manner that is functionally centralized, while preserving the scalability, flexibility, and fault tolerance that are characteristic of self-organizing systems.

Bio
Marco Dorigo received the Ph.D. degree in electronic engineering in 1992 from Politecnico di Milano, Milan, Italy. From 1992 to 1993, he was a Research Fellow at the International Computer Science Institute, Berkeley, CA. In 1993, he was a NATO-CNR Fellow, and from 1994 to 1996, a Marie Curie Fellow. Since 1996, he has been a tenured Researcher of the FNRS, the Belgian National Funds for Scientific Research, and co-director of IRIDIA, the artificial intelligence laboratory of the ULB. His current research interests include swarm intelligence, swarm robotics, and metaheuristics for discrete optimization. He is the Founding Editor of Swarm Intelligence, and an Associate Editor or member of the Editorial Boards of many journals on computational intelligence and adaptive systems. Dr. Dorigo is a Fellow of the AAAI, EurAI, and IEEE. He was awarded numerous international prizes among which the Marie Curie Excellence Award in 2003, delivered by Philippe Busquin, European Commissioner for Research, Science and Innovation; the five-year scientific prize du F.R.S.-FNRS in 2005, delivered by Albert II, King of Belgium; the Cajastur International Prize for Soft Computing „Mamdani Prize” in 2007, delivered by Prof. Lofti Zadeh; the IEEE Frank Rosenblatt Award in 2015, and the IEEE Evolutionary Computation Pioneer Award in 2016. He is also the recipient of an ERC Advanced Grant awarded by the European Research Council.

Prof. Christian Napoli, La Sapienza Roma

Title: Modeling Time and Structure: Modern Challenges and Old Tricks for Trustworthy AI Systems

Trustworthy artificial intelligence systems must operate reliably over time and support human interaction in real-world conditions. However, many modern data-driven approaches focus on static performance metrics and treat data as collections of independent samples, overlooking temporal organization and structural dependencies that are intrinsic to signals and dynamic processes.

This talk discusses why modeling time and structure is essential for building trustworthy AI systems, and why several effective solutions rely on principles that are well known but often neglected in current practice. By analyzing common failure modes in signal-based and evolving data, the talk shows how ignoring temporal and structural information leads to unstable behavior, reduced robustness, and loss of user trust.

Rather than proposing new architectures, the focus is on design principles: representation choices, inductive bias, and simple modeling strategies that improve stability, predictability, and interpretability over time. Examples from real systems illustrate how combining modern learning techniques with these “old tricks” can lead to AI systems that are not only accurate, but also reliable and usable in human-centered scenarios. The talk concludes by outlining open challenges for trustworthy AI operating in non-static environments.

Bio
TBA.

TBA

Title: TBA

Abstract:
TBA

Bio
TBA

PP-RAI 2026

PP-RAI 2025