Plenary Speaker of 2026

Professor Emeritus

Budapest University of Technology and Economics, Hungary

Speech Title: Fuzzy signatures, situational maps and some application examples

Abstract: In many real life problems connected with engineering and other applied sciences, a series of often vague, noisy or ill-defined features, which in most cases may be grouped into several subsets, where the components are related closer to each other. Sometimes even sub-subsets and sub-sub-subsets nay be identified within these subsets, by expert knowledge, or, more often, by statistical interdependence. Such structures may be represented by either a rooted tree graph, or an iteratively nested “vector”, which may have sub-vectors as components. The connections among branches of any subtree must be also defined within the structure. This happens by the application of fuzzy aggregation operators, which are assigned to all intermediate node within the tree graph (including the root). These aggregations may be also selected, and their respective parameters estimated by expert knowledge, but it is more adequate to obtain them by Machine Learning, such as applying evolutionary optimization algorithms, or artificial neural networks – whenever a sufficient real data base is available where the resulting evaluation and all corresponding inputs (the leaf memberships) are known. The final evaluation happens by executing all the aggregations level by level, starting at the leaves of the graph, and obtaining the final fuzzy value (membership degree) in the root. Such constructions are called Fuzzy Signatures (FSigs). FSigs are extensions of the original concepts of Fuzzy Sets, and of Vector Valued Fuzzy Sets as well.

Thus Fuzzy Signatures are complex, hierarchically structured descriptors, which allow the presence of various types of uncertainty, especially, vagueness, but also statistical features. Because of the intention of applying this model class to real life problems, it is expected that they are suitable for manipulations (aggregations among themselves, other operations, comparisons, etc.). A really interesting case occurs when the available data are not homogeneous, and thus, their respective actual tree structures are also not entirely identical, because e.g. some components are missing or completely covered by noise. This is the case, e.g., when a doctor describes the health condition of some patients by a FSigs, and the symptoms are not homogeneously available, e.g., the body temperature of one patient is known for several days before the model is set up, and in another case, only recently. The presentation will also give an introduction to the definitions and basic operations in connection with FSigs.

A special extension of the idea of FSig is given by the Fuzzy Situational Maps (FSM) where one or more sub-trees may represent matrices of two (or even more) dimensions with more or less fixed spatial structure, not unlikely as in hidden maps. Zoom in and Zoom out operations combined with proper fuzzy aggregations help to increase or decrease the detail view of a given part of the area described by the FSM, and at the same time, allow partial complexity reduction by shrinking/compressing the information into higher hierarchical levels – a very convenient tool for describing the positions and movements of objects, such as mobile robots, obstacles, etc. Three dimensional situational maps may be of practical use when civil engineering constructions are examined and tested by 3D networks of sensors, such as acoustic sensors measuring the speed of sound going through masonry, and thus establishing information of hidden irregularities within the masonry.

A couple of possible application areas of FSM will be presented, namely the fuzzy communication and collaboration of intelligent mobile robots (the problem originally set by the LIFE research team lead by the late Prof. T. Terano in Yokohama, in the 1990-s).
Then a couple of civil engineering and architecture related problems will be introduced, the description of the condition of historical residential buildings in Budapest (Hungary), modelled by FSigs, and then the experiments carried out by a team at the University of Pécs concerning masonry strength estimation will be shown. This latter application lead to a new method of optimizing the aggregations within the FSig model, by applying various aggregation families and parameters will be shown.

 

Biography: Dr. h.c. László T. Kóczy has Ph.D. and D.Sc. degrees and is a member of the Polish Academy of Sciences, a Fellow of the International Fuzzy Systems Association, and multiple other scientific societies, he received the Fuzzy Pioneer Award from the IEEE Computational Intelligence Society, etc.
He Is Professor Emeritus at Budapest University of Technology and Economics, he was Dean and President of the University Ph.D. Council at Széchenyi István University (Győr), where he works now part time as a Professor Emeritus, and he is now Full Professor at John von Neumann University (Kecskemét), and is Dean’s Senior Advisor at Pécs University of Sciences.
He has been a visiting professor in various countries in four continents, among others, in Japan, Korea, Australia, Italy and Austria.
He published close to 900 scientific books, journal and conference papers, and received about 10,000 citations to them.
He received the Officer and Commander Crosses of the Order of Hungary.

 

 

Keynote Speakers of 2026

Member of the Academia Europaea
IEEE Fellow, EAI Fellow, AAIA Fellow
ACM Distinguished Scientist
ACM Distinguished Speaker
AIIA (InternationaI AI Industry Alliance) president

TU Wien, Austria

Speech Title: Active Inference for Distributed Intelligence in the Computing Continuum

Abstract: Modern distributed systems must operate under uncertainty, across environments, infrastructures, and applications that vary widely.
Within the Computing Continuum (IoT–Edge–Fog–Cloud), applying neuroscience-inspired principles and mechanisms may help us build more flexible solutions that can generalize across diverse settings.
Intriguing hypotheses in neuroscience propose that many brain functions in humans and animals arise from a small number of powerful principles.
If these hypotheses hold, they could offer deep insight into how humans and animals cope with unpredictable events—and even support imagination.
In this talk, we explore how Active Inference, alongside established design principles for modern distributed systems—such as elasticity, predictive equilibrium, and antifragility—can enable Distributed Intelligence across the Computing Continuum.

Biography: Schahram Dustdar is a Full Professor of Computer Science at TU Wien, where he leads the Distributed Systems Group, and he is also affiliated as an ICREA research professor at Universitat Pompeu Fabra
(UPF) in Barcelona. (https://dustdar.prof) He is widely known for pushing the frontier of elastic, dependable cloud-to-edge systems and the Computing Continuum, turning cutting-edge research into practical foundations for modern distributed intelligence. An IEEE Fellow, ACM Distinguished Scientist/Speaker, and member of Academia Europaea, he’s a sought-after keynote speaker recognized for shaping how large-scale systems adapt, scale, and stay resilient in the real world.

 

 

 

SMIEEE
Lead Data and AI Ethics Group / Lead Machine Intelligence Group

Manchester Metropolitan University, UK

Speech Title: From Uncertainty to Responsibility: The Road to Responsible AI

Abstract: One of the enduring challenges in artificial intelligence has been the development of methods capable of supporting decision-making under conditions of uncertainty, ambiguity, and incomplete information. This lecture traces a research journey from early work on reasoning under uncertainty to contemporary challenges surrounding the responsible design, deployment, and governance of AI systems. Drawing on experiences across multiple generations of AI technologies and application domains, the talk explores how increasing real-world deployment exposed important limitations in purely technical approaches to AI, prompting a shift towards broader questions of transparency, accountability, fairness, human oversight, and public trust. Drawing on these lessons, the lecture explores what Responsible AI means in practice and how organisations can balance increasingly capable AI systems with human expertise, judgement, and contextual understanding. Particular attention is given to trust calibration, accountability, transparency, and the distribution of responsibility within Human-AI Teams as AI systems become more autonomous and agentic. The lecture concludes by considering the role of public engagement, governance, and responsible innovation in developing human-centred AI ecosystems that maintain meaningful human control while delivering societal value and sustaining public confidence in the age of AI.

Biography: Keeley Crockett SMIEEE SFHEA is a Professor in Computational Intelligence at Manchester Metropolitan University She has over 27 years’ experience of research and development in Ethical and responsible AI), computational intelligence algorithms and applications, including adaptive psychological profiling, fuzzy systems, semantic similarity, and dialogue systems. Keeley was a steering committee member for the Policy Connect and UK Parliamentary Inquiry on Skills in the Age of AI. She is one of the five EPSRC Public Engagement Champions and was PI on the EPSRC “PEAs in Pods: Co-production of community based public engagement for data and AI research.” Keeley was one on the Founders of the People Panel for AI, funded by The Alan Turing Institute and adopted in 2024 Manchester City Council. She is the co-academic AI lead at MMU for the UKRI funded Centre for Digital Innovation and is working on several Innovate UK KTPs. Keeley was a collaborator on the AI Playbook for the UK Government and in Sept-2025 co-chaired the AI in Government and Academia Summit. She is part of the International Agentic AI Safety Experts Focus Group which includes the development of the Guidelines for Agentic AI Safety Volume 1&2 (2024, 2025). Keeley Chairs the UKRI’s AI & Robotics Strategic Advisory Team. She founded the IEEE Technical committee SHIELD (2022-2024) and currently Chairs the new IEEE CIS AI Ethics Education and Awareness Taskforce.