[EAAS] Emerging IT/OT-Security concepts from Brownfield to Future Industrial Environments

Proposers: Prof. Mike Barth, M.Sc. Marwin Madsen,

Abstract: Emerging cyber-criminal threads targeting industrial production systems result in the need to advance the state of IT/OT-Security. This necessity is highlighted by the increasing importance of compliance with security directives such as the Network and Information Security Directive (NIS2) and the Cyber Resilience Act (CRA). Therefore, this special session addresses recent advances in industrial automation regarding their security impacts. By fostering a collaborative environment, this session will provide a platform for discussing the latest advancements in IT/OTSecurity research, sharing developed methods, toolchains and techniques as well as best practices, and exploring new strategies to enhance IT/OT security in compliance with the upcoming regulations. This is not limited to pure technical aspects as understanding the role of human behaviour and organizational culture in security practices is crucial for developing comprehensive security strategies. Topics of interest include, but are not limited to, threat detection and response, risk management, secure communication protocols, and the role of artificial intelligence in enhancing IT/OT security.

[ICONS] Fairness and Equity in Engineering: a Control Systems Perspective

Proposers: Dr. Saverio Bolognani, Marina Ceccon, Matteo Cederle

Abstract: The proposed special session aims to delve into the overlooked intersection of fairness/equity and control systems engineering. While fairness has been extensively studied by the artificial intelligence (AI) community in various sectors (from recruitment/hiring to recommendation systems), its application to control systems remains relatively unexplored, even though in recent years there has been a growing recognition of the importance of fairness and equity in several other engineering areas. Given the pervasive presence of control systems in engineering domains, this special session seeks to bridge this gap by bringing together experts, researchers, and practitioners to examine how control systems design and implementation can influence the fairness and equity of engineering solutions. Ensuring fairness in control systems is essential for promoting equitable distribution of resources, opportunities, and outcomes. By addressing this topic, the special session aims to not only raise awareness but also foster discussion and catalyze research efforts towards developing more socially responsible engineering systems and to promote the role of control systems approaches in this area.
Furthermore, it will also explore how fairness and equity specifications can be incorporated into control system design and how they can be encoded in mathematical certificates and guarantees.
The special session is designed to engage researchers, PhD students and practitioners in the fields of control systems engineering, and, secondarily, of related engineering disciplines. Additionally, it welcomes researchers in the AI community and researchers interested in the ethical and societal implications of engineering solutions. By bringing together diverse perspectives, the session aims to stimulate interdisciplinary dialogue and collaboration. We envision to attract control systems experts working in several application domains where fairness and equity could have a central role, such as (but not limited to):
- resource allocation in smart grids;
- traffic management systems;
- healthcare delivery systems;
- autonomous vehicles and mobility services

[TA] Resilience in Distributed Control and Optimization

Proposers: Dr. David Angeli, Dr. Alessandro Casavola, Dr. Sabato Manfredi

Abstract: In an era where complex systems increasingly rely on distributed control and optimization, ensuring resilience against uncertainties, disturbances, and adversarial influences is paramount. This special session focuses on novel methodologies that enhance the robustness, adaptability, and security of distributed control frameworks deployed in cyber-physical systems, smart grids, multi-agent networks, and large-scale optimization problems. Resilience in distributed control entails the ability to maintain system performance despite failures, delays, and attacks. Recent advancements in networked control, decentralized optimization, and fault-tolerant algorithms offer promising solutions to mitigate risks and ensure system stability. This session invites contributions that address theoretical foundations, algorithmic innovations, and realworld applications that improve resilience in dynamic and uncertain environments. Key topics include but are not limited to: resilient consensus algorithms, distributed optimization under adversarial conditions, Vickrey–Clarke–Groves (VCG) mechanisms for resilient optimization and control, fault-tolerant multi-agent control, resilient Model Predictive Control and Command Governor schemes, learning-based resilient control strategies, and security-aware distributed decisionmaking. The session aims to bring together researchers and practitioners to discuss cutting-edge developments and foster collaborations toward building more resilient and intelligent distributed control systems.

[ICONS] Advances in modelling, learning, and control of complex supply and service chains

Proposers: Dr. Nikolaos Athanasopoulos, Prof. Seán McLoone

Abstract: This special session focuses on advances in modelling, learning, analysis, estimation and decision-making algorithms for complex cyber-physical systems. Of particular interest are dynamical systems found in supply chains and service chains. We invite contributions on challenges found in planning, scheduling and resource allocation, stemming from the adoption of new paradigms within manufacturing, industrial applications, (predictive) maintenance, logistics, transportation networks, Industry 4.0/5.0 applications, including Manufacturing as a Service, Services Virtualisation, and Intent-based modelling, estimation and control. The session aims to feature contributions demonstrating cutting- edge research from both industry and academia using tools from AI and reinforcement learning, discrete-event systems, hybrid systems, set based methods, and computational intelligence tools

[TA] Control and AI approaches for sustainable Industry 5.0

Proposers: Pal Varga, Germar Schneider, Markus Tauber, Mike Holenderski

Abstract: This special session focuses on the control and AI approaches for interoperable IoT systems and sustainable Industry 5.0, addressing the challenges of seamless connectivity, intelligent automation, and resource-efficient industrial operations. As Industry 5.0 evolves toward a human-centric, resilient, and sustainable framework, the integration of interoperable IoT systems with advanced control and AI techniques is essential for optimizing industrial processes. This session aims to bring together researchers and industry practitioners to explore innovative solutions that enhance real-time data exchange, decentralized decision-making, and adaptive control strategies in IoT-enabled smart manufacturing and sustainable industrial environments.
Topics of interest include, but are not limited to, edge AI for IoT-driven process optimization, digital twins for real-time industrial control, AI-enhanced predictive maintenance, secure and scalable IoT architectures, interoperability standards for cyber-physical systems, sustainable manufacturing, and the human aspects of AI in the industrial workplace. Contributions that showcase novel methodologies and practical applications in smart factories, autonomous production lines, and next-generation industrial networks are particularly encouraged. This session will serve as a platform to discuss emerging trends and breakthroughs that leverage AI and control strategies to enable fully interoperable, adaptive, and sustainable Industry 5.0 ecosystems.

[TA] Advances in Intelligent Control Theory and Applications for Linear and Nonlinear Systems

Proposers: Dr. Yanzhi Wu

Abstract: This special session focuses on cutting-edge advancements in Intelligent Control Theory and Applications for Linear and Nonlinear Systems, a pivotal area driving innovation in automation, trains. With the growing complexity of modern engineering challenges, the integration of intelligent control methodologies—such as adaptive control, data- driven optimization, and neural networks—has become essential for addressing dynamic, uncertain, and nonlinear behaviors in real-world systems. This session will showcase interdisciplinary research bridging theoretical rigor with practical implementation, covering topics including (but not limited to): novel control frameworks for nonlinear dynamics; AI-enhanced control algorithms (e.g., reinforcement learning, deep predictive control);robustness and adaptability in heterogeneous environments;applications in autonomous systems.

[ICONS] Advancements and Applications in Reinforcement Learning

Proposers: Dc. Alberto Sinigaglia, Dc. Davide Sartor, Dc. Niccolò Turcato, Prof. Lucian Busoniu

Abstract: This special session invites contributions highlighting recent advances and diverse applications of Reinforcement Learning (RL), a dynamic and growing field within artificial intelligence that empowers agents to learn optimal behaviors through interactions with their environments. As RL methodologies evolve, they continue to be successfully adopted across numerous domains, including robotics, autonomous systems, gaming, simulation, industrial processes, healthcare, and beyond.
We welcome a broad range of research and practical papers addressing innovative RL algorithms and techniques, including, but not limited to, model-based and model-free approaches, deep reinforcement learning, hierarchical RL, and multi-agent reinforcement learning (MARL). Submissions demonstrating novel or impactful applications of RL in various domains such as robotic manipulation, autonomous navigation, simulated environments, gaming, healthcare systems, industrial automation, manufacturing, process control, and energy management are strongly encouraged.
The objective of this session is to foster an inclusive platform for knowledge exchange among researchers and industry practitioners, emphasizing recent achievements, emerging trends, and future opportunities within the expansive landscape of Reinforcement Learning and its versatile applications.

[ICONS] Computational Systems: Interpretability, Optimization, and Emerging Challenges

Proposers: Rokas Gipiškis, Dr. Davide Dalle Pezze, Eng. Valentina Zaccaria

Abstract: This special session focuses on recent research in interpretability, optimization, and other related emerging challenges in computational systems. As AI-driven technologies become more complex and widely deployed, ensuring their transparency, efficiency, and reliability is critical. This session aims to bring together researchers and practitioners to explore advancements in explainable AI, optimization techniques, and the challenges posed by evolving computational paradigms. This session seeks to bring together experts from academia and industry to highlight both foundational research and practical solutions to the most pressing challenges in computational systems. We seek innovatve solutons that leverage explainable AI, efficient learning algorithms, decision-making under uncertainty, ethical and regulatory considerations, and applications in fields such as healthcare, manufacturing, and autonomous systems.
Topics of interest include, but are not limited to:
- Interpretability and Explainability: Techniques for understanding AI models and improving the transparency in decision-making processes.
- Optimization: Novel algorithms and approaches for optimizing systems.
- Emergent Paradigms: Advancements beyond classic supervised classification, such as Domain Adaptation, Continual Learning, Semi-Supervised Learning, Unsupervised Learning, and Meta-Learning.
- Efficient Learning Algorithms: Developing algorithms that are resource-efficient, scalable, and capable of handling large datasets and real-time processing.
- Decision-Making Under Uncertainty: Approaches to decision-making in complex, uncertain environments, including reinforcement learning, probabilistic modeling, and decision theory.
- Human-AI Interaction: Improving the collaboration between humans and AI systems, including interface design, user experience, and trust-building in autonomous systems.
- Privacy-Preserving Machine Learning: Techniques that ensure data privacy in machine learning models, including the Federated Learning framework.

[ICONS] Intelligent Control and Optimization of Wind and Marine Energy Systems

Proposers: Dr. Matilde Santos, Dr. J. Enrique Sierra-García, Dr. Payam Aboutalebi

Abstract: To achieve energy goals with renewable alternatives, control of wind and marine energy systems remains a challenge for engineers. The main difficulties from the control point of view come from the fact that it must meet several objectives simultaneously, along with the complex and non-linear dynamics of the marine energy systems and varying environmental conditions. Addressing the problem of control of these marine energy devices also involves dealing with other related issues such as model identification and optimization of their operation. This has led to the inclusion of artificial intelligence techniques in control strategies in order to provide effective solutions.
The objective of this special session is to bring together research papers on different aspects of wind and offshore energy systems, mainly focused on control and optimization of their operation, carried out by means of intelligent computing. The session will discuss different proposals for these energy systems, including the combination of conventional and intelligent control techniques, heuristic optimization, intelligent approaches for modeling energy converters, and potential applications to physical wind and offshore energy systems.
Topics include, but are not limited to, artificial intelligence techniques applied to modeling, identification and optimization of wind and marine energy conversion systems; intelligent control of marine renewable energy systems; digital twins targeting control of onshore and offshore wind turbines; intelligent experimental and numerical/simulation approaches; operation and maintenance on wind turbines and marine energy platforms, etc

[TA] Control and Data-Driven Approaches in IoT and Smart City Systems

Proposers: Eng. Francesco Borsatti, Dr. Federico Chiariotti, Dr. Marco Fabris

Abstract: The convergence of Edge Computing and IoT is reshaping modern control systems, enabling real-time decisionmaking, adaptive automation, and distributed intelligence across industrial and cyber-physical applications. However, the distributed and networked nature of edge-enabled IoT systems introduces new challenges related to computational constraints, latency, decentralized control, and the efficient management of heterogeneous data streams in resource-constrained environments.
This special session aims to bring together researchers and practitioners working on both methodological and application-driven contributions at the intersection of edge computing, networked control, and data-driven approaches. We seek innovative solutions that leverage control theory, machine learning, optimization, and reinforcement learning to address the challenges of distributed decision-making, resource allocation, and autonomous operation in resource-constrained Edge-IoT ecosystems.
Topics of interest include, but are not limited to:
- Edge-native control strategies for IoT-enabled systems, addressing latency constraints and real-time adaptation
- Reinforcement Learning (RL) and Multi-Agent RL (MARL) for distributed decision-making in resource constrained Edge-IoT environments
- Age of Information (AoI) and Value of Information (VoI) approaches for remote estimation and control tasks in IoT networks
- Networked and distributed control systems, including delay-aware and event-triggered control architectures
- Hybrid model-based and data-driven approaches, integrating physics-based models with AI-driven adaptation
- Edge AI and federated learning for distributed processing and optimization in resource-limited Edge-IoT networks
- Distributed access schemes for networked control and estimation systems
- Digital twins and predictive maintenance, leveraging IoT data for real-time system modeling and anomaly detection
- Joint communication and control strategies under energy, computational, and latency constraints
- Security and privacy-aware control in Edge-IoT networks, ensuring resilience against cyber threats and data integrity concerns
- Scalability and interoperability challenges in resource-constrained Edge-IoT deployments
- Edge-cloud coordination, optimizing computational distribution between edge devices and centralized infrastructure
We particularly welcome contributions that explore the interplay between networked control, distributed intelligence, and Edge-IoT architectures, emphasizing how edge computing influences the performance, efficiency, and reliability of IoT-driven control solutions. Reinforcement Learning-based approaches, especially in decentralized and multi-agent settings, are of special interest, as they provide promising frameworks for adaptive, self-learning, and scalable control strategies in computationally and energy-constrained environments.
The session encourages both theoretical advancements and real-world case studies demonstrating the impact of control, AI, and edge-native methodologies in distributed, resource-constrained Edge-IoT applications.

[ICONS] Learning-Based and Advanced Control Strategies for Physical Systems

Proposers: Dr. Dalla Libera Alberto, Prof. Masero Rubio Eva, Prof. Bruschetta Mattia, Prof. Rampazzo Mirco, Prof. Carli Ruggero

Abstract: In recent years, the interplay between machine learning, identification, and control has led to the derivation of several advanced frameworks for identification and control, including data-driven control, learning-based Model Predictive Control (MPC), and Reinforcement Learning. The application of these methodologies to physical systems is particularly compelling in industrial contexts, offering opportunities to enhance control performance and boost energy efficiency. This special session aims to highlight recent advancements in learning-based and advanced control strategies applied to physical systems across various domains, such as robotics, industrial automation, and energy-efficient systems. The featured research emphasizes both theoretical innovations and practical applications, showcasing the effectiveness and transformative potential of these algorithms. Contributions span a diverse range of topics, from energy optimization in industrial systems to sophisticated motion planning and control strategies for robotic manipulators. This session will serve as a platform for discussing cutting-edge developments, exchanging ideas, and fostering collaboration among experts in control systems, optimization, and robotics.

[ICONS] Machine Learning and Control approaches for Anomaly Detection and Predictive Maintenance

Proposers: Dr. Lucas Brito, Prof. Olga Finkm, Dr. Davide Frizzo, Prof. Chia-Yu Hsu

Abstract: The increasing complexity and interconnectedness of modern systems demand sophisticated approaches for ensuring reliability, safety, and efficiency. Among diagnostic and prognostic strategies, Anomaly Detection (AD), Predictive Maintenance (PdM) and Prescriptive Maintenance (RxM) are critical components in achieving these goals, enabling early identification of potential failures and proactive maintenance strategies. The aim of this Special Session is to explore innovative techniques for Anomaly Detection, Predictive Maintenance and Prescriptive Maintenance in various application domains. We aim to bring together researchers and practitioners from both the machine learning and control communities to discuss novel methodologies and practical implementations. We seek contributions that leverage the strengths of both fields to develop robust, accurate, and efficient solutions for detecting anomalies, predicting future failures, and optimizing maintenance schedules. Topics of interest include, but are not limited to: - Machine learning techniques for anomaly detection in complex systems, including supervised, unsupervised, and semi-supervised methods.
- Control-theoretic approaches for anomaly detection, such as fault detection. fault identification and fault isolation.
- Hybrid approaches combining machine learning and control theory for enhanced AD, PdM and RxM performances.
- Data driven Predictive Maintenance strategies based on Machine Learning and Deep Learning for Remaining Useful Life (RUL) estimation.
- Reinforcement Learning based approaches for optimal maintenance decision-making and scheduling.
- Explainable AI (XAI) for AD, PdM and RxM.
- Robustness and uncertainty quantification in machine learning and control-based AD, PdM and RxM.
- Integration of domain expertise and physics-based models in anomaly detection and maintenance strategies.
We particularly encourage contributions that address the challenges of dealing with high-dimensional data, unlabelled or partially labelled data, noisy measurements, and evolving system dynamics. Methodologies that incorporate domain knowledge and physics-based models are also of great interest. The session welcomes both theoretical advancements and practical case studies demonstrating the effectiveness of machine learning and control approaches in real-world anomaly detection and proactive maintenance applications. We are especially interested in works that show how the combination of these two fields can lead to more effective and reliable solutions than either field alone.

[ICONS] Human-adaptive approaches for collaborative robotics in Industry 5.0

Proposers: Dr. Giulio Giacomuzzo, Dr. Matteo Terreran, Dr. Mattia Guidolin

Abstract: This special session focuses on the ongoing transformation in the industry, particularly in the realm of Human-Robot Collaboration, which is reshaping modern manufacturing. The shift from isolated, highly specialized machines to flexible, autonomous systems enables safer and more efficient collaboration between humans and robots in shared workspaces. This evolution promises to enhance productivity and adaptability but also introduces new challenges related to human monitoring, motion prediction, intention recognition, and adaptive control. Nonetheless, these challenges are at the core of the novel Industry 5.0 paradigm, which proposes a human-centric approach to industrial automation. This special session aims to explore theoretical advancements and practical applications of human-robot collaboration in modern industry, focusing on human-centered interaction and seamless teamwork in shared workspaces. Topics include, but are not limited to, human perception, human action and intention recognition, human motion prediction, robot learning from human demonstration, dynamic task allocation, adaptive robot control, human-aware task and motion planning, and trust assessment.

[ICONS] Humans in the Loop: Bridging Data Science and Automatic Control with Human-Centric Applications

Proposers: Prof. Damiano Varagnolo, Prof. Simone Del Favero

Abstract: The integration of human factors into data-driven applications and into closed-loop systems is becoming increasingly critical across diverse domains, from healthcare to workplace safety. This invited session aims to explore how data science, machine learning and closed-loop control can be leveraged to address challenges where human behavior, physiology, and decision-making play a central role. The session will feature cutting-edge research that highlights the interplay between data-driven and closed-loop methodologies and human-centric applications. The contributions gathered for this session will span a broad variety of case studies, such as:
- Gait Analysis in Children. A contribution will investigate the performance of various data imputation algorithms in addressing missing data in pediatric gait analysis, a critical step in ensuring accurate diagnostics and personalized interventions.
- Prediction of Female Orgasms from Physiological Signals. A contribution will investigate which features from time-series data from ECG, galvanic skin response, and breathing patterns are the most powerful in predicting physiological orgasms.
- Injury Risk Detection for Workers Using Smartwatch Data. A contribution will explore how wearable technology can monitor physical exertion and detect injury risks in real-time, offering a proactive approach to workplace safety.
- Human Errors in Type 1 Diabetes Therapy. A contribution will evaluate, though a sensitivity analysis, the impact of human errors in both automated (closed-loop) and manual (open-loop) diabetes management systems, providing insights into the design of more robust and human-aware therapeutic solutions.
Together, these contributions offer a broad view on the importance of accounting for human factors into data-driven and closed-loop systems for human-centric applications, ensuring that technological advancements are not only accurate but also aligned with human needs and behaviors. This session will foster interdisciplinary discussions and inspire innovative solutions at the intersection of data science and human-centric applications.

[TA] Digital Transformation of Health and Social Infrastructure of Smart Cities and Communities

Proposers: Prof. David Bogataj, Prof. Alenka Temeljotov Salaj

Abstract: This special session at the IFAC Conference highlights the transformative potential of emerging technologies to enhance care, rehabilitation, and quality of life for older adults. It emphasises innovative applications of AI, robotics, ambient intelligence, virtual reality (VR), and tracking technologies. The session synthesises recent literature and presents key debates from diverse interdisciplinary perspectives. Given Europe's rapidly aging population, the demand for effective health and social care services is significantly increasing. Older adults often face declining functional abilities, necessitating the development of adaptive living and work environments. To address these challenges, public and private sectors must innovate through new policies, technological advancements, and methodologies explicitly tailored for the aging population. Social innovations play a pivotal role, introducing novel products, services, and collaborative models to meet societal needs and enhance social interaction, thereby improving overall well-being. Consequently, there is significant growth in developing and testing advanced ICT solutions, including e-Health, m-Health, e-Care, m-Care, Ambient Assisted Living (AAL), and Ambient Intelligence technologies. These innovations empower elderly individuals to maintain independence, substantially contributing to the formation of inclusive social infrastructure within Smart Cities and aging communities. They also drive economic growth and sustainability within the burgeoning Silver Economy.