Kyri Baker

⚡ New Research Publication! Excited to share that our paper, "Solving Large-Scale Unit Commitment Problems by Physics-Informed Graph Learning," is now published in IEEE Transactions on Power Systems! 🎉 In this work, we introduce a physics-informed graph learning framework that revolutionizes the way unit commitment (UC) problems—critical to electricity market clearing—are solved. By embedding power system physics and network topology directly into a graph neural network, our method delivers lightning-fast and scalable UC predictions without sacrificing accuracy. 💥 What’s even more exciting: our approach significantly outperforms the industry-leading solver Gurobi in both speed and scalability, especially on large-scale UC problems with complex system constraints. This breakthrough brings us one step closer to enabling real-time, AI-powered market clearing and grid operations under growing uncertainty and renewable penetration. Here is the link to the publication https://lnkd.in/gnQanFt4 #PowerSystems #UnitCommitment #ElectricityMarkets #GraphNeuralNetworks #AI #SmartGrid #IEEE #Optimization #EnergyInnovation #Gurobi #GridModernization #MachineLearning

91

1 Comment

📢 Preprint alerts! Not one, not two...but *four* manuscripts submitted from TIME Lab in May!🔥🔥🔥🔥 - Julian Najera and Hao Chen from Scott Howard's lab in University of Notre Dame, Department of Electrical Engineering led a story exploring how #mechanics alters #metabolism in the peritumoral microenvironment of #breastcancer using #FLIM. - Maksym Zarodniuk and Anna Wenninger from Christopher Patzke's lab in University of Notre Dame - College of Science led an effort to uncover the cellular and molecular mechanisms underlying #neuron damage and #neuroinflammation from #braintumor #mechanics in mice and humans. - Alice Burchett continued her investigation of #mechanoimmunology of #myeloid cells and found that #mechanics promote a pro-tumor pathological state in #macrophages. - Golnaz Asaadi Tehrani found that #epigenetic modifiers can sensitize 3D #spheroid models of #glioblastoma to chemotherapy. Links in the comments below!

111

7 Comments

The Spring issue of IEEE Solid-State Circuits Magazine is out! https://lnkd.in/gxbjNtKw This issue is dedicated to Integrated Circuits for AI and ML applications, and was guest edited by Marian Verhelst and Charlotte Frenkel. It features four excellent articles that highlight core challenges and methodologies in AI hardware acceleration (see https://lnkd.in/gayteYmV for more detail). The feature articles are: 1) "Design and Optimization of Efficient Digital Machine Learning Accelerators" by Wei Tang, Sung-Gun Cho, Jie-Fang Zhang, and Zhengya Zhang examines AI acceleration through optimized computational kernels, such as matrix–matrix multiplication, leveraging architectures like SIMD and systolic arrays (10.1109/MSSC.2025.3549361). 2) "SRAM- and eDRAM-Based Compute-in-Memory Designs, Accelerators, and Evaluation Frameworks," by Yifan He, Xiaofeng Hu, Hongyang Jia, and Jae-sun Seo addresses in-memory computing (IMC) as a means to reduce data movement overhead, covering both analog and digital IMC approaches (https://lnkd.in/gHiukWKx). 3) "Open Source Heterogeneous SoCs for Artificial Intelligence," by Francesco Conti, Angelo Garofalo, Davide Rossi, Giuseppe Tagliavini, and Luca Benini explores the integration of heterogeneous systems on chip, demonstrating how open source platforms such as PULP support scalable AI hardware development (https://lnkd.in/gAQgK-8d). 4) " Sparsity-Aware Hardware," by Man Shi, Adrian Kneip, Nicolas Chauvaux, Jiacong Sun, Charlotte Frenkel, and Marian Verhelst discusses sparsity-aware hardware design, illustrating how structured and unstructured sparsity can be leveraged for significant performance improvements using an adapted roofline model (https://lnkd.in/gv8eTyvd). The issue also features regular columns: 1. Behzad Razavi’s “The Analog Mind” column (https://lnkd.in/gxKyqZr7) continues the series on CMOS inverter applications 2. Chris Mangelsdorf’s “Shop Talk: What You Didn’t Learn in School” column picks up where the previous issue left off, continuing the discussion on Miller compensation in "Miller’s Wrong Half-Plane Zero" (https://lnkd.in/gngTrywV). This time, the focus is on the right half-plane zero and its potential to degrade stability if not properly addressed. 3. Farhana S.’s “Circuits From a Systems Perspective” traces the historical evolution of wireless charging in everyday devices (https://lnkd.in/gnVhqUyF). Lot's of interesting articles to read! Enjoy! #ieeesscc #analog #ieee #ai #ml #solidstatecircuits #hardwareacceleration #semiconductors #inmemorycomputing

260

3 Comments

If self-driving car companies do not develop principled and repeatable testing methodologies for end-to-end-learning, this is not going to be the last time we hear about threats to pedestrians, cyclists, people on scooters and other drivers https://lnkd.in/eU6pdBxH

109

16 Comments

We launched a "quantum tapeout" course at Berkeley! The course covers fundamentals and computer-aided design of superconducting quantum circuits. Students will submit their designs in 10 weeks and get to test the chips by the end of the term. The course materials are available here: https://lnkd.in/gQ3M7dyb We hope the references, notes, HWs, and (simulation) lab assignments will be self-contained enough to introduce interested engineers and physicists to the field. Suggestions and feedback welcomed! Supported by Challenge Institute for Quantum Computation (CIQC).

821

25 Comments

Ending 2025 on a high note. My colleague Professor Kris Davis and I have collaborated over the past several years to elucidate the role of atomic layer deposition (ALD) in the development of carrier-selective contacts for silicon (Si) photovoltaics (PV). The outcome of this effort is a comprehensive review with insights from 300+ publications, highlighting key trends and technological advances in Si PV enabled by ALD-based carrier-selective contacts. We propose the standardization of PV test structures to facilitate meaningful comparison and improved data consistency across the field. Our hope is that this article will serve as a valuable resource for both ALD practitioners and Si PV device and process engineers, and help guide future innovations in silicon photovoltaic technologies. Thanks to Dr. Chien-Hsuan Chen for his stupendous effort on this article! Thanks to our funding agencies #NationalScienceFoundation (#NSF) and #DepartmentofEnergy (#DOE), and #UniversityofCentralFlorida (#UCF) for supporting our work.

104

4 Comments

I am excited to share HIENet, a foundation model for materials science that achieves SOTA performance and efficiency. HIENet combines E(3)-invariant and O(3)-equivariant message passing layers to achieve superior performance while more than doubling the inference speed compared to existing foundation models. We find that adding invariant message passing layers retains the impressive performance scaling of equivariant models without the massive computational overhead of extra tensor product operations. Additionally, we use gradient-based force and stress calculations to ensure HIENet satisfies important physical constraints such as equivariance and force field conservation. We find that satisfying these physical constraints, beyond just force/stress equivariance, is important for downstream applications. In our preprint, we show that HIENet achieves impressive results on common machine learning benchmarks, and more importantly we further demonstrate its utility on a variety of downstream materials science tasks such as calculating phonon band structures, elasticity tensors, bulk moduli, alloy phase diagrams, and molecular dynamics simulations. Our preprint and model checkpoints are available to use now with training code coming soon. Preprint: https://lnkd.in/g6Q-iznY Code: https://lnkd.in/gqjZGG-p Authors: Keqiang Yan, Montgomery Bohde, Andrii Kryvenko, Ziyu Xiang, Kaiji Zhao, Siya Zhu, Saagar Kolachina, Doğuhan Sar��türk, Jianwen Xie, Raymundo Arroyave, Xiaoning Qian, Xiaofeng Qian, Shuiwang Ji

141

4 Comments

Excited to share Alexander Barajas-Ritchie's first journal paper of his Ph.D. in our research group. We propose and develop an integration platform for Wave Energy Converter models and power system simulation software (both commercial and open source). With robust and modular software, researchers and practitioners are better equipped to run studies at scale and facilitate reproducibility with different benchmarks and tools. Open access manuscript in SoftwareX. Good work Alex! https://lnkd.in/g7FaWHSY

27

Now published in Chemical Reviews: a review of the various techniques for benchmarking quantum gate operations, including a detailed description of our new Deterministic Benchmarking (DB) method. This work was led by Dr. Vinay Tripathi (formerly Daniel Lidar's PhD student, now at IBM) and Dr. Daria Kowsari (postdoc in my group), with important contributions from Kumar Saurav (current Lidar PhD student) and Dr. Haimeng Zhang (former Lidar and Levenson-Falk PhD student, now at IBM). We hope this is a useful resource for the community! https://lnkd.in/gTUpYUzn

64

Battery charging protocol design is an intricate craft. While the literature offers a forest of physics-based and data-informed models, the real question is: what actually makes it into commercial devices, and why? What mechanistic tradeoffs guide the engineers and scientists who turn theory into something that quietly works in your phone or electric vehicle? In this Advanced Energy Materials article, we examine the charging protocols used in consumer electronics and EVs. We map out where different products converge or diverge, and we probe the design logic that shaped the features we see today. This team effort is led by three graduate students, Wenlong Li, Ayrton Yanyachi, and Junyi Xia, at the Walker Department of Mechanical Engineering, The University of Texas at Austin and the Texas Materials Institute, The University of Texas at Austin. The collaboration among my The University of Texas at Austin colleagues, Dr. Yixian Wang, Prof. David Mitlin, Prof. O.A. (DK) Ezekoye, and Prof. Junmin Wang, really demonstrates a strong momentum that we are building. We are also thankful to the close collaboration with Dr. Saravanan Kuppan of Rivian, Dr. Jigang Zhou of General Motors, Prof. Piero Pianetta of SLAC National Accelerator Laboratory, Prof. Kejie Zhao of Purdue University, Dr. Yanfei Li and Prof. Jie Xiao of Pacific Northwest National Laboratory. https://lnkd.in/gSJFGNxJ

123

1 Comment