DISC- Summer School 2025

TU Delft DISC_online

Cyber-physical systems, where feedback loops are closed over communication channels, are ubiquitous in many domains, including utility, transportation, robotics and industrial systems. In these settings, the physical component refers to the entity or environment that needs to be controlled and the cyber component is the computing and communication technologies that integrate the system. The finite capacity of communication technologies inadvertently introduces a bottleneck in guaranteeing the same stability and performance as a system with infinite information flow between the cyber and physical components. This situation becomes predominant in growing multi-agent networks, where the network capacity is shared among the agents. This summer school aims to provide the fundamentals of estimation and control under various facets of limitations in information flow between the cyber and physical components of the system, including the data rate and modeling for control systems. Deterministic and stochastic approaches for linear, nonlinear and switched systems will be covered.

Target audience: PhD students, postdocs, and researchers interested in the field.
This summer school will be taught by international and local experts.

Summer School 2025 – in retrospective

When the organizers of the DISC summer school 2025 first conceived of the idea, we wanted a 3.5 day course on fundamentals and advances in the estimation and control of control systems today. Eventually, we narrowed it down to addressing two of the most important aspects in recent years: (i) communication constraints and vulnerabilities; and (ii) poor model knowledge. Thus, the 2025 edition of the DISC summer school was born, with the title “Estimation and Control under Coarse Information”. We have the privilege and pleasure of engaging leading experts in addressing both aspects on the following topics:

Emulation-based control and estimation

Dragan Nesic (University of Melbourne) on ‘Emulation-based controller design using time and event-triggered schemes’.
Michelle Chong (Eindhoven University of Technology) on “Emulation-based observer design and Secure state estimation”

Minimum data rates

Girish Nair (University of Melbourne) on ‘Minimum data rates for imitating a linear systems’
Sasha Pogromsky (Eindhoven University of Technology) on ‘Minimum data rates for remote state estimation’
Daniel Liberzon (University of Illinois Urbana-Champaign) on ‘Minimum data rates for estimation, localization and related problems’

Data informativity

Kanat Camlibel (Groningen University) on ‘Data informativity for data-driven modeling, analysis and control’
Henk van Waarde (Groningen University) on ‘Experiment design for data-driven modeling, analysis and control’

Neural networks

Bahman Gharesifard (Queens University) on ‘The interplay between estimation, control and training of neural networks’

Through lectures and exercises, 18 participants from Dutch universities and abroad were exposed to the fundamentals and cutting-edge research in the aforementioned topics. We also got to know each other’s work better as participants were encouraged to bring a poster on their work, which provided a basis for discussion throughout the duration of the school and dedicated poster interactive sessions.

As organizers of the 2025 edition of the DISC summer school, we are truly thankful to DISC for their generous support, to the external speakers for their commitment and boundless enthusiasm, and to the participants for their dedication. We are also grateful to Maaike, the DISC secretary who has been relentless in supporting us in the organization of the school. We had tremendous fun putting together the program and teaching at the school. Concurrently, we have also learnt a lot from each other. We look forward to continuing the discussions and friendships beyond the summer school.

Regards,

The organizers of the 2025 DISC Summer School:
Michelle Chong, Sasha Pogromsky, Henk van Waarde and Kanat Camlibel