WS-02 Alternative Representations
for Control: Neuromorphic, Data-driven,
and Moment-based methods

Abstract 

 Projection operators appear in many modern control architectures, enforcing stability, safety, and performance by modifying nominal controllers through projection onto constraint sets. Examples include projections enforcing stability through Control Lyapunov Functions (CLFs), safety through Control Barrier Functions (CBFs), and performance through tailored input–output constraints. This talk presents a perspective on projected dynamics in control, connecting Projected Dynamical Systems, CBF-based safety filters, and regularization techniques reminiscent of anti-windup control. We highlight their common structure, discuss key differences, and illustrate their role in applications ranging from online feedback optimization (FO) to high-performance mechatronic motion systems used in semiconductor industries.

Bio 

 W.P.M.H. Heemels received M.Sc. (mathematics) and Ph.D. (EE, control theory) degrees (summa cum laude) from the Eindhoven University of Technology (TU/e) in 1995 and 1999, respectively. From 2000 to 2004, he was with the Electrical Engineering Department, TU/e, as an assistant professor, and from 2004 to 2006 with TNO-Embedded Systems Institute as a Research Fellow. Since 2006, he has been with the Department of Mechanical Engineering, TU/e, where he is currently a Full Professor and Vice-Dean. He held visiting professor positions at ETH, Switzerland (2001), UCSB, USA (2008) and University of Lorraine, France (2020). He is a Fellow of IEEE and IFAC, and was the chair of the IFAC Technical Committee on Networked Systems (2017-2023). He served/s on the editorial boards of Automatica, Nonlinear Analysis: Hybrid Systems (NAHS), Annual Reviews in Control, and IEEE Transactions on Automatic Control, and is the Editor-in-Chief of NAHS as of 2023. He was a recipient of a personal VICI grant awarded by NWO (Dutch Research Council) and an ERC Advanced Grant. He was the recipient of the 2019 IEEE L-CSS Outstanding Paper Award and the Automatica Paper Prize 2020-2022. He was elected for the IEEE-CSS Board of Governors (2021-2023). His current research includes hybrid and cyber-physical systems, networked, neuromorphic, and event-triggered control systems and model predictive control and their applications.

Abstract 

 The adjacent fields of reinforcement learning (RL) and adaptive control share the same objectives, yet they are separated by a wide cultural gap. In this presentation, I attempt to bridge this gap for the linear quadratic regulator (LQR) problem, which serves as a cornerstone and the benchmark for both fields. I begin by discussing different learning pipelines, including direct and indirect (model-based) approaches, as well as episodic and online (adaptive) approaches. Despite the extensive literature spanning several decades, numerous problems remain unsolved. For instance, RL methods are seldom concerned with closed-loop stability The adjacent fields of reinforcement learning (RL) and adaptive control share the same objectives, yet they are separated by a wide cultural gap. In this presentation, I attempt to bridge this gap for the linear quadratic regulator (LQR) problem, which serves as a cornerstone and the benchmark for both fields. I begin by discussing different learning pipelines, including direct and indirect (model-based) approaches, as well as episodic and online (adaptive) approaches. Despite the extensive literature spanning several decades, numerous problems remain unsolved. For instance, RL methods are seldom concerned with closed-loop stability certificates or efficient implementations, while the adaptive control community has dedicated minimal effort to optimality. We address the data-driven LQR problem in an adaptive setting, which entails online recursive algorithms and closed-loop data, and we seek both algorithmic as well as closed-loop certificates. Our approach encompasses different variations of policy gradient methods, employs a novel covariance parameterization, and leverages different kind of regularizations. Finally, all our theoretical results are validated through simulations and experiments in diverse domains, demonstrating the computational and sample efficiency of our method.

Bio 

 Florian Dörfler is a Professor at the Automatic Control Laboratory at ETH Zürich. He received his Ph.D. degree in Mechanical Engineering from the University of California at Santa Barbara in 2013, and a Diplom degree in Engineering Cybernetics from the University of Stuttgart in 2008. From 2013 to 2014 he was an Assistant Professor at the University of California Los Angeles. He has been serving as the Associate Head of the ETH Zürich Department of Information Technology and Electrical Engineering from 2021 until 2022. His research interests are centered around automatic control, system theory, optimization, and learning. His particular foci are on network systems, data-driven settings, and applications to power systems. He is a recipient of the 2025 Rössler Prize, the highest scientific award at ETH Zürich across all disciplines, as well as the distinguished career awards by IFAC (Manfred Thoma Medal 2020) and EUCA (European Control Award 2020). He and his team received best paper distinctions in the top venues of control, machine learning, power systems, power electronics, circuits and systems. They were recipients of the 2011 O. Hugo Schuck Best Paper Award, the 2012-2014 Automatica Best Paper Award, the 2016 IEEE Circuits and Systems Guillemin-Cauer Best Paper Award, the 2022 IEEE Transactions on Power Electronics Prize Paper Award, the 2024 Control Systems Magazine Outstanding Paper Award, and multiple Best PhD thesis awards at UC Santa Barbara and ETH Zürich. They were further winners or finalists for Best Student Paper awards at the European Control Conference (2013, 2019), the American Control Conference (2010,2016,2024), the Conference on Decision and Control (2020), the PES General Meeting (2020), the PES PowerTech Conference (2017,2025), the International Conference on Intelligent Transportation Systems (2021), the IEEE CSS Swiss Chapter Young Author Best Journal Paper Award (2022,2024,2025), the IFAC Conferences on Nonlinear Model Predictive Control (2024) and Cyber-Physical-Human Systems (2024), and NeurIPS Oral (2024). He is currently serving on the council of the European Control Association and as a senior editor of Automatica.

Abstract 

TBD

Bio 

 Anders Rantzer (Fellow, IEEE) received the Ph.D. degree in optimization and systems theory from KTH Stockholm, Stockholm, Sweden, in 1991. He is a Professor of automatic control with Lund University, Lund, Sweden. He was a postdoctoral at IMA, University of Minnesota, 1992/93. He was a visiting associate faculty member at Caltech and 2015/2016 he was Taylor Family Distinguished Visiting Professor, University of Minnesota. He is a member of the Royal Swedish Academy of Engineering Sciences, Royal Physiographic Society in Lund and former Chairman of the Swedish Scientific Council for Natural and Engineering Sciences. His research interests include modeling, analysis and synthesis of control systems, with particular attention to scalability, adaptation, and applications in energy networks.

Abstract 

 We study how to predict the future evolution of unknown dynamical systems from past windows. We first construct a library of prototypical trajectories (templates), generated by different systems. We then combine these templates to build a new repertoire of sample trajectories that can successfully predict the future even when the observed past windows do not belong to any system in the original library. This demonstrates a form of generalization across dynamical behaviors. 

Bio 

 Claudio De Persis is a professor with the Engineering and Technology Institute Groningen, University of Groningen, the Netherlands, since 2011. He received the Laurea degree in electronic engineering in 1996 and the PhD degree in systems engineering in 2000, both from the University of Rome "La Sapienza", Italy. Before joining the University of Groningen, he held postdoctoral and faculty positions at Washington University in St. Louis, Yale University, the University of Rome "La Sapienza" and Twente University.

Abstract 

 Understanding who influences whom in a nonlinear multi-agent network is increasingly important in data-driven control, distributed autonomy, and network science. In this talk, we present a framework for efficient inference of weighted interaction graphs in diffusively coupled nonlinear networks using steady-state probing experiments. The approach exploits the network’s steady-state input-output map, rather than its transient dynamics, enabling recovery of topology and edge weights from a small number of carefully designed perturbations. This viewpoint is particularly useful beyond the linear setting, where existing graph recovery methods are more limited, especially for networks with non-smooth or discontinuous local dynamics. Under global convergence assumptions that can be certified using passivity-based tools, the proposed method applies to a broad class of nonlinear systems and yields an algorithm with sub-cubic computational complexity; in noiseless linear time-invariant networks, the recovery is exact. We will also discuss robustness to disturbances and measurement errors, the effect of limited probing access, and a matching complexity-theoretic lower bound showing that the method is essentially time-optimal. 

Bio 

 Daniel Zelazo is a Full Professor in the Stephen B. Klein Faculty of Aerospace Engineering at the Technion—Israel Institute of Technology, where he leads the Cooperative Networks and Controls Lab (ConNeCt). He received his B.Sc. and M.Eng. degrees from MIT and his Ph.D. from the University of Washington, and was a postdoctoral research associate at the University of Stuttgart before joining the Technion in 2012. He currently serves as an Associate Editor for the IEEE Transactions on Control of Network Systems, and has previously served as an Associate Editor for IEEE Control Systems Letters and the International Journal of Robust and Nonlinear Control. His research focuses on systems and control theory, multi-agent and networked dynamical systems, optimization, and graph-theoretic methods for complex interconnected systems.

Abstract 

TBD

Bio 

 Bayu Jayawardhana received the bachelor degree in electrical and electronics engineering from the Institut Teknologi Bandung, Bandung, Indonesia, in 2000, the M.Eng. degree in electrical and electronics engineering from the Nanyang Technological University, Singapore, in 2003, and the Ph.D. degree in electrical and electronics engineering from Imperial College London, London, U.K., in 2006. He is currently the Scientific Director of Dutch Institute of Systems and Control, the Scientific Director of Engineering and Technology Institute Groningen, Faculty of Science and Engineering, University of Groningen and Director of Groningen Engineering Center, University of Groningen. His research interests include the analysis of nonlinear systems, systems with hysteresis, optomechatronics, multi-robot systems and systems biology. He is a vice-chair of publications in the IFAC Technical Committee on Nonlinear Control Systems and is a fellow of the Netherlands Academy of Engineering. 

Abstract 

TBD

Bio 

 Francesco Bullo is a Distinguished Professor of Mechanical Engineering at the University of California, Santa Barbara, CA, USA. He received the Laurea degree in Electrical Engineering from the University of Padova, Italy, in 1994, and the Ph.D. degree in Control and Dynamical Systems from the California Institute of Technology, Pasadena, CA, in 1998. His research interests include contraction theory, network systems, neural networks, and mathematical neuroscience. He is the author or coauthor of Geometric Control of Mechanical Systems (Springer, 2004), Distributed Control of Robotic Networks (Princeton, 2009), Lectures on Network Systems (KDP, v1.7, 2024), and Contraction Theory for Dynamical Systems (KDP, v1.3 2026). He served as IEEE CSS President and SIAG CST Chair. He is a Fellow of ASME, IEEE, IFAC, NetSci, and SIAM.

Abstract 

 Moment matching is a model reduction technique that allows the construction of reduced-order models preserving specific moments. These moments are associated with the steady-state output response of the system to be reduced, interconnected in an open-loop fashion with a signal generator. Model matching thus relies on strong stability properties and on the availability of a model of the signal generator.

 To relax these requirements, in the first part of the talk we introduce a data-driven procedure for computing reduced-order models from input-output data generated by an agnostic signal generator. The moments are directly identified from the output of the system to be reduced, and the resulting reduced-order models achieve asymptotic matching.

 To relax the stability requirements, in the second part of the talk we revisit the notion of moment matching by introducing the concept of closed-loop interpolation. This notion relies upon the construction of a novel class of signal generators, which are feedback-interconnected with the plant to be reduced, and allows the definition of moments and models for unstable systems. The existence of a family of models that parameterise all, possibly unstable, systems achieving moment matching in a closed-loop fashion is also presented.

Bio 

 Alessandro Astolfi (IFAC Fellow, FIEEE, Academia Europaea, ITATEC) received the Laurea in Electronic Engineering from the University of Rome La Sapienza, Italy, in 1991; the M.Sc. degree in Information Theory and the Ph.D. degree with Medal of Honor with a thesis on discontinuous stabilization of holonomic systems from ETH-Zürich, Zürich, Switzerland, in 1995; and the Ph.D. degree for his work on nonlinear robust control from the University of Rome "La Sapienza," Rome, in 1996.

 In 1992, he joined ETH-Zürich as a research associate. Since 1996, he has been with the Electrical and Electronic Engineering Department, Imperial College London, London, U.K., where he is currently Professor of Nonlinear Control Theory. From 2022 till 2025 he was College Consul for the Faculty of Engineering and Business School and from 2010 to 2022, he was the Head of the Control and Power Group at Imperial College London. From 1998 to 2003, he was an Associate Professor with the Department of Electronics and Information, Politecnico of Milano, Milano, Italy. Since 2005, he has also been a Professor with the Dipartimento di Ingegneria Civile e Ingegneria Informatica, University of Rome Tor Vergata, Rome. 

 His research interests include mathematical control theory and control applications, with special emphasis for the problems of discontinuous stabilization, robust and adaptive control, observer design, optimal control, game theory, and model reduction. 

 Dr. Astolfi was the recipient of the IEEE CSS A. Ruberti Young Researcher Prize (2007); the IEEE RAS Googol Best New Application Paper Award (2009); the IEEE CSS George S. Axelby Outstanding Paper Award (2012); the Automatica Best Paper Award (2017); and the IEEE Transactions on Control Systems Technology Outstanding Paper Award (2023). He is a "Distinguished Member" of the IEEE CSS, IFAC Fellow, IET Fellow, and Member of the Academia Europaea. He is the recipient of the Institute of Measurement and Control Sir Harold Hartley Medal for "Outstanding contributions to the technology of measurement and control".

 He was Editor-in-Chief of the IEEE Transactions on Automatic Control (2018 -- 2025) and Chair of the IEEE CSS Conference Editorial Board (2010–2017). He is Vice Chair of the IFAC Technical Board (2020–2026).

Abstract 

 In this presentation, we introduce an alternative perspective for analyzing and synthesizing the input-output behavior of a system by confining ourselves to inputs generated by a dynamical system, often referred to as an exosystem. While this might initially appear to be a restrictive assumption, we illustrate through our recent preliminary results on linear time-invariant (LTI) systems that this might actually provide a structured pathway toward a broader mathematical framework.

 In particular, we first show that steady-state assignment problems such as output regulation can be elegantly resolved purely out of the Sylvester equation. This allows the perspective that the fundamental roles of subsystems in interactions are nothing but the virtual exosystems and the moment transfer operators that they represent. Meanwhile, the moment is a data-dependent concept, depending specifically on the choice of its associated exosystem. In the second part, we illustrate that moments are also capable of what resembles direct data-driven control, demonstrating how moment-based approaches can drastically simplify controller synthesis. Finally, in the third part, we introduce a practical relaxation of Willems' fundamental lemma, revealing that inputs generated by exosystems are in fact sufficiently rich while also being efficient enough to facilitate both analysis and design. 

 We conclude by discussing the potential for expanding this framework to nonlinear dynamics via the nonlinear counterpart of the Sylvester equation, i.e., the invariance equation, with an illustration of the future research roadmaps toward neuromorphic control.

Bio 

 Jin Gyu Lee received the B.S. and Ph.D. degrees from Seoul National University, Korea, and held the post-doc position at University of Cambridge, United Kingdom till 2021. He held another post-doc position at Imperial College London, United Kingdom in 2022. He joined Inria, Lille, France, in 2022. Since 2024, he has been with Seoul National University as an assistant professor. His research interests include multi-agent systems, observer design, secure control systems, nonlinear oscillators, distributed optimization, adaptive control, neuronal networks, and neuromorphic engineering.