We share ideas, build community, and support learning through teaching, workshops, tutorials, and seminars.
Teaching
Courses, lecture materials, student projects, and other educational resources from the lab.
EN.560.654 (01)
Introduction to Machine Learning and Control for Building Energy Systems
This course introduces the principles of building energy optimization with a focus on the modeling and control of HVAC systems. The curriculum covers HVAC fundamentals, thermodynamics, and heat transfer, progressing to dynamic systems, control theory, and optimization techniques. Key topics include an introduction to system identification, machine learning, and optimal control applied to energy-efficient building management. The course concludes with hands-on coding assignments focusing on implementation of machine learning and control techniques for optimizing the energy efficiency of buildings using high-fidelity simulation frameworks.
EN.560.652 (01)
Scientific Machine Learning for Modeling, Optimization, and Control of Dynamical Systems
This course offers a scientific machine learning (SciML) approach to the modeling, optimization, and control of dynamical systems. Students will learn to systematically integrate physics-based models and constraints into deep learning architectures, and to leverage data-driven methods for accelerating the solution of large-scale optimization and optimal control problems. Key topics include physics-informed neural networks, learning to optimize, neural differential equations, neural operators, and differentiable control. The course also examines real-world applications of these emerging SciML techniques in domains such as building energy management, networked dynamical systems, and power systems. Emphasis will be placed on practical, hands-on coding exercises and project-based assessments to reinforce theoretical concepts through implementation.
Workshops
Focused programmes that bring together ideas from control, optimization, scientific computing, and machine learning.
Physics Meets Data: Learning, Optimization, and Control
A full-day workshop at the 23rd IFAC World Congress, co-organized by Ján Drgoňa, Thomas Beckers, Sandra Hirche, and Rolf Findeisen. The programme combined invited talks with hands-on sessions in physics-informed machine learning.
AI for Modeling, Control, and Optimization of Partial Differential Equations
ACC 2026 workshop spanning control theory, machine learning, and applications of AI to PDE systems. Ján presented “Learning to Control PDEs with Differentiable Predictive Control and Neural Operators.”
Physics-Informed Learning for Cyber-Physical Systems
A CPS-IoT Week workshop on using physics-informed machine learning, formal methods, and differentiable programming for scalable and safe modeling, optimization, and control.
Tutorials
Open lectures, code, and practical materials for learning scientific machine learning methods.
Scientific Machine Learning for Modeling, Optimization, and Control
Organized by Thomas Beckers, Truong X. Nghiem, and Ján Drgoňa, this tutorial develops a unified differentiable-programming view of learning to model, learning to optimize, and learning to control.
Seminars & Invited Talks
Invited presentations by SOLARIS members, arranged in reverse chronological order by year.
17th World Congress on Computational Mechanics
Invited talk by Ján Drgoňa at WCCM–ECCOMAS 2026.
Scientific Machine Learning for PDE Control and Mixed-Integer Optimization
CCPS seminar by Ján Drgoňa on differentiable predictive control and learning-to-optimize methods for scalable real-time decision-making.
Invited Talk at TU Dortmund
Invited research talk by Ján Drgoňa.
Invited Talk at Ruhr University Bochum
Invited research talk by Ján Drgoňa.
Applied Mathematics Seminar
Invited seminar by Ján Drgoňa.
Bose Research Seminar
Invited online seminar by Ján Drgoňa.
Accelerating Underground Pumped Hydro Energy Storage Scheduling with Decision-Focused Learning
Invited talk by Honghui Zheng in the “Machine Learning for Global Optimization” session.