Workshop:
AI for Control Systems: Learning, Reasoning, and Adaptation in Safety-Critical
Dynamical Systems
Artificial intelligence is rapidly reshaping the design and operation of control systems across robotics, autonomous vehicles, power and energy systems, and broader cyber-physical infrastructure. Yet integrating AI into feedback-driven dynamical systems remains a fundamental challenge: how can learning, reasoning, and large-scale models be incorporated into control loops while preserving stability, robustness, and operational safety in safety-critical environments?
This workshop brings together researchers and practitioners working at the intersection of machine learning and control systems to address this challenge from both theoretical and application-driven perspectives. We focus on learning-enabled feedback control, safe and constrained reinforcement learning, hybrid model-based and data-driven architectures, adaptive control augmented with machine learning, and verification and certification of learning-integrated controllers.
Scope and topics:
The workshop is structured around two major safety-critical application domains: (i) intelligent power and energy systems, where reliability, grid stability, and uncertainty mitigation under high DER penetration are essential; and (ii) robotics and autonomous systems, where adaptive and learning-based controllers must ensure safe interaction, disturbance rejection, and real-time decision-making under physical constraints. While these domains differ in scale and infrastructure, both demand AI-enabled control frameworks that are reliable, interpretable, certifiable, and deployable under real-world uncertainty.
Topics of Interest Include:
Session 1: Safe and Reliable AI for Power and Energy Systems- Advanced machine learning for power and energy systems
- Safe reinforcement learning in power system operation and control
- Uncertainty mitigation with extensive Distributed Energy Resource (DER) integration
- Sustainable and resilient energy systems enabled by AI
- Multi-agent system-based management and coordination in smart grids
- Explainable AI (XAI) applications for power system monitoring and decision-making
- Human-in-the-loop ML applications for grid supervision and operational support
- Learning-based fault detection and predictive maintenance in power infrastructure
- Robust and trustworthy ML methods for energy market and grid stability analysis
- Learning-enabled feedback control systems for robotics and autonomous systems
- Adaptive and nonlinear control augmented with machine learning components
- Safe and constrained reinforcement learning for robotic control
- Stability-aware and Lyapunov-based learning control methods
- Hybrid model-based and data-driven control architectures
- Online system identification and adaptive learning in robotic systems
- Learning-based disturbance estimation and rejection
- Vision-based and multimodal learning for robotic manipulation and navigation
- Verification, validation, and certification of learning-enabled control systems
- Real-time and embedded ML deployment for safety-critical control applications
Chairs:
- Van-Hai Bui:
vhbui@umich.edu
- Guilherme V. Hollweg:
hollweg@umich.edu
- Karishma Patnaik:
kpatnaik@umich.edu
- We will accept (1) short papers reporting preliminary results or (2) Full research papers
- All submissions will undergo peer review by the workshop technical committee. Accepted papers will be presented as oral presentations.
- At least one author of an accepted paper must register and present.
- The accepted papers will be available online in our workshop webpage.
- CMT Submission Site
- Select the track: Workshop: AI for Control Systems: Learning, Reasoning, and Adaptation in Safety-Critical Dynamical Systems
- Submission Deadline: June 20, 2026
- Notification of Acceptance: July 10, 2026
Dr. Van-Hai Bui, IEEE Senior Member, received his B.E. degree in Electrical Engineering from
Hanoi
University of Science and Technology, Vietnam, in 2013, and his Ph.D. degree in Electrical
Engineering
from Incheon National University, South Korea, in 2020. From 2022 to 2023, he served as an Assistant
Professor in the Department of Electrical Engineering at the State University of New York (SUNY)
Maritime College, USA. Since August 2023, he has been an Assistant Professor in the Department of
Electrical and Computer Engineering at the University of Michigan-Dearborn, USA. His research
interests
include energy management systems, applications of machine learning in smart grids, and operation &
control of power and energy systems. Dr. Bui's research has been sponsored by the U.S. National
Science
Foundation (NSF), the Michigan Institute for Data and AI in Society (MIDAS), Ford Motor Company, and
multiple local industrial partners.
Dr. Guilherme V. Hollwegreceived the B.Sc., M.Sc., and Ph.D. degrees in Electrical Engineering from the Federal University of Santa Maria (UFSM), Santa Maria, Brazil, in 2016, 2019, and 2021, respectively. He is currently an Assistant Professor with the Department of Electrical and Computer Engineering at the University of Michigan–Dearborn, Dearborn, MI, USA. He serves as the IEEE-HKN faculty advisor for the UM-Dearborn Theta Tau Chapter. His research interests include adaptive and nonlinear control theory, optimization-based control design, and their applications in power electronics, motor drives, renewable energy systems, microgrids, and robotics. His work particularly focuses on robust and adaptive control strategies for power converters, electric machines, and cyber-physical systems.
Dr. Karishma Patnaik is an Assistant Professor in the Department of Electrical and Computer Engineering at the University of Michigan-Dearborn. She has a Ph.D. in Systems Engineering from Arizona State University (ASU) in 2023 and M.Tech. in Systems and Control Engineering from IIT Bombay in 2016. She is the recipient of several awards, including the ASME Rising Stars in Mechanical Engineering (2022), the ASU Graduate College Completion Fellowship (2023), and the Outstanding Postdoctoral Scholar from ASU Faculty Women of Color Caucus (2025). At UM-Dearborn, she leads the NICHE Robotics Lab, developing control algorithms for robot-environment physical interaction in safety-critical tasks such as manufacturing, maintenance and agriculture. Her research interests include contact modeling and adaptive control of nonlinear systems, interaction control, hybrid dynamical systems, optimal control, generative planning, and decision-making.
Paper Submission Instructions
All papers will be double-blind reviewed and must present original work.
Papers submitted for reviewing should conform to IEEE specifications. Manuscript templates can be downloaded from:
Keydates
Registration
In order for your paper to be published in the proceedings you must register to the conference.
Workshop Structure
Coffee and Breakfast - 8:30am-9am
Session 1.1: Invited Talks (Morning) 9am-12pm, 3 invited talks 45 min each + 10 min Q & A + 5 min Setup Invited talks on AI-integrated control systems, Learning in feedback loops, Safety and verification in AI-based controllers
Session 1.2: Panel Discussion, 12:00-12:30pm Theme: “How do we certify AI-enabled control in industry?”
Lunch Break - 12:30pm-1:30pm
Session 2.1: Paper Presentations - “Safe and Reliable AI for Power and Energy Systems”, 1:30pm-3pm, 6 papers, 15 min each (8min presentation and 5min Q&A + 2min setup)
Coffee Break - 3pm-3:15pm
Session 2.2: Paper Presentations - “Learning-Enabled Control for Safety-Critical Robotics and Autonomous Systems”, 3:15pm-4:45, 6 papers, 15 min each (8min presentation and 5min Q&A + 2min setup)
Session 2.3: Best Workshop Paper Awards, 4:45-5pm
ICMLA'26