Hackathon
Pietro at the Tesseract Hackathon 2026
Pietro took part in the Tesseract Hackathon 2026 with FIDE—Fiber-Informed Differentiable Experimental Design—working on optimal experimental design for systems observed through aggregate measurements.
Scalable Optimization, Learning, And Robustness for Intelligent Systems
ExploreJohns Hopkins University
SOLARIS LAB
Scalable Optimization, Learning, And Robustness for Intelligent Systems
Innovating at the intersection of Physics, AI, and Engineering.
Science can amuse and fascinate us all, but it is engineering that changes the world.
Partners in discovery
Sponsors & Affiliations
We gratefully acknowledge the institutions and programs that support and connect our work.
From the lab
Latest News
Paper
Eight SOLARIS Papers Accepted at CDC 2026
SOLARIS will have a strong presence at the 65th IEEE Conference on Decision and Control in Honolulu. Congratulations to Liang, Pietro, Honghui, and everyone across the lab and our collaborator network behind this outstanding collective achievement.
View the eight-paper lineup
- Learning to Control PDEs with Differentiable Predictive Control and Time-Integrated Neural Operators
- SCORE: Statistical Certification of Regions of Attraction via Extreme Value Theory
- Fixed-time-stable ODE Representation of LassoAlso accepted as an IEEE Control Systems Letters paper.
- Stability Enhanced Gaussian Process Variational Autoencoders
- Mixed-Integer Nonlinear Differentiable Predictive Control for Underground Pumped Hydro Energy Storage Systems
- Safe Receding Horizon Mixed-Integer Differentiable Predictive Control for Degradation-Aware Battery Dispatch
- Koopman Meets Discrete-Time Control Barrier Functions: A Linear Model Predictive Control Framework
- Polynomial Parametric Koopman Operators for Stochastic MPC
We gratefully acknowledge support from the U.S. DOE Office of Science, ASCR programme, through the SciDAC Institute LEADS, and from the Ralph O’Connor Sustainable Energy Institute at Johns Hopkins University.
Paper
Honghui's First Journal Paper
Congratulations to Honghui on his first journal paper, “Accelerating Underground Pumped Hydro Energy Storage Scheduling with Decision-Focused Learning,” now published in IEEE Transactions on Sustainable Energy.
Grant
DOE Genesis Mission Award
A Johns Hopkins team led by Ján received a $750,000 Phase I award for “From Chip to Chiller: Verifiable Edge AI Agents for Data Center Thermal Management.” The project will develop real-time AI agents that reduce cooling energy while maintaining reliable data-center operation.
Paper
Three SOLARIS Papers at IFAC 2026
Three papers from the group were presented at the 23rd IFAC World Congress in Busan, spanning differentiable control for data-center cooling and mixed-integer optimal control, and time-certified Koopman NMPC. Congratulations to Liang and all our collaborators.
Workshop
Physics Meets Data Workshop at IFAC 2026
Ján co-organized the full-day workshop “Physics Meets Data: Learning, Optimization, and Control” at the IFAC World Congress in Busan, combining invited talks with hands-on sessions in physics-informed machine learning.
Grant
NSF Funding for AI-Driven Autonomous Laboratories
Ján is part of the Johns Hopkins AIMD-Net team building a network of AI-driven autonomous laboratories for faster materials discovery. The four-year programme is supported by a $20 million National Science Foundation grant.
Talk
Four Invited Talks Across Germany
Ján shared SOLARIS research in a four-stop invited-talk series at Ruhr University Bochum, TU Dortmund, TU Darmstadt, and the 17th World Congress on Computational Mechanics in Munich.
View the itinerary
- 17th World Congress on Computational Mechanics
- Technical University DarmstadtScientific Machine Learning for PDE Control and Mixed-Integer Optimization
- Technical University Dortmund
- Ruhr University Bochum
Tutorial
L4DC Scientific Machine Learning Tutorial Now Online
Recordings, slides, code, and hands-on notebooks are now available for “Scientific Machine Learning for Modeling, Optimization, and Control,” organized by Thomas, Truong, and Ján at L4DC 2026.
Talk
Poster Presentation
Pietro presented the poster `CINOC: Cardinality-Invariant Neural Operator Policies for Scalable PDE Control` at ICML 2026.
Talk
Tutorial
Ján presented the tutorial `Scientific Machine Learning for Modeling, Optimization, and Control` at L4DC 2026.
Team
Welcome, Nima
Welcome to Nima, who has joined the lab.
Talk
Invited Talk
Ján gave an invited lecture at the Workshop on AI for Modeling, Control, and Optimization of Partial Differential Equations during the American Control Conference (ACC) in New Orleans.
Talk
Poster Presentation
Honghui presented the research poster `NOIR: A MINLP Solver using Neural Overparameterization for Integer Representation` at the University of Connecticut - Stamford.
Talk
Invited Talk
Ján was a speaker for the Bose research seminar series (online).
Talk
Workshop Organization
Ján organized a workshop on Physics-Informed Learning for Modeling, Optimization, and Control of Cyber-Physical Systems at the Cyber-Physical Systems and Internet-of-Things (CPS-IoT) Week in France.
Talk
Invited Talk
Ján presented an invited seminar to the Applied Math Department at the University of California, Santa Barbara.
Paper
Liang's Paper Accepted to L-CSS
Congratulations to Liang on the acceptance of `Fixed-time-stable ODE Representation of Lasso` to L-CSS.
Talk
Invited Talk
Honghui gave an invited talk on `Accelerating Underground Pumped Hydro Energy Storage Scheduling with Decision-Focused Learning` at the INFORMS Optimization Society Conference.
Team
Welcome, Leandro
Welcome to Leandro, who has joined the lab from EPFL.
Award
1st Place Tesseract Hackathon 2025
Congratulations to Honghui, Dibakar, and Pietro for winning first place in the Tesseract Hackathon 2025.
Team
Welcome, Liang
Welcome to Liang, who has joined the lab.
Team
Welcome, Parv
Welcome to Parv, who has joined the lab.
Team
Welcome, Honghui and Pietro
Welcome to Honghui and Pietro, who have joined the lab.