Data-centric and learning-based methods have pervaded all areas of science, engineering, technology, and society at large, including the field of control systems. Data-driven control has emerged as a dominant theme, intersecting with developments in machine learning, operations research, and practical engineering applications. Historically rooted in adaptive control, system identification, and variations of adaptive dynamic programming, data-driven control builds on a rich legacy of techniques from our field blended with recent advances in machine learning, optimization, and statistics. Recent advances in algorithms and realtime computation have expanded the reach and capabilities of these approaches. The control community not only adopts but also contributes to the broader AI topic by providing theoretical rigor and interpretability to black-box models. This Open Invited Track seeks to embrace a broad definition of data-driven control, encouraging contributions in the wide area of data-driven decision making, that span data-driven control design, machine learning theory for dynamic model building and control, and practical applications. The session aims to highlight the critical role of data-driven methods in shaping the future of control systems, fostering a diverse and collaborative research community that reflects the growing relevance of this interdisciplinary domain.

• Data-driven control design
• Control oriented modeling and uncertainty quantification based on data conditions, exploration-exploitation tradeoffs
• Analysis of data-driven control methods, such as safe and reliable integration of learning-based components into control loops: stability guarantees, robustness under distribution shifts, and formal verification of data-driven controllers
• Fundamental limitations, e.g., concerning sample complexity or excitation
• Control-theoretic perspectives on machine learning algorithms
• Applications data-driven control methods to real-world systems
• Data-driven system analysis
• Numerical methods for data-driven control

For the submission code of the OIT please check here

• Florian Dörfler, ETH Zürich
• Alessandro Chiuso, University of Padova
• Julian Berberich, University of Stuttgart
• Valentina Breschi. TU Eindhoven
• Timm Faulwasser, Hamburg University of Technology
• Simone Formentin, Politecnico di Milano
• Mircea Lazar, TU Eindhoven
• Guanru Pan, Hamburg University of Technology
• Yoshihiko Susuki, Kyoto University
• Henk van Waarde, University of Groningen
• Karl Worthmann, TU Ilmenau