Strengthen your skills through “Insight to Impact” data visualization and science communication workshops
PRI’s Insight to Impact workshop series equips researchers with advanced tools and strategies for effective science communication at any stage of their careers. The series features experts covering topics ranging from mastering media relations to transforming complex science into impactful narratives.
These workshops will be held on select Wednesdays, usually from 3–4 pm. Details are below.
Upcoming Workshops
April 7 — Scientific Machine Learning – Towards Instant Modeling, Design, and Digital Twins
Seid Koric, Ph.D., will present his team’s research in Scientific Machine Learning (SciML), focusing on new variants of Deep Operator Neural Networks (DeepONets). These advanced neural architectures learn mappings from parametrically varying inputs—such as loads, boundary conditions, materials, or geometries—to full solution fields of Partial Differential Equations (PDEs) governing physical systems. Trained on limited data from classical numerical simulations, DeepONets can accurately predict nonlinear multiphysics 2D and 3D solution fields without retraining, achieving speedups of up to 10,000× over traditional supercomputer-based methods. This enables near-instant forward evaluations within inverse design, optimization, sensitivity analysis, uncertainty quantification, and real-time digital twin applications across engineering, medical, and natural sciences.
Speaker Biography
Koric received his B.S. in Mechanical Engineering from the University of Sarajevo, Bosnia and Herzegovina (1993), and his M.S. in Aerospace Engineering (1999) and Ph.D. in Mechanical Engineering (2006) from the University of Illinois Urbana-Champaign. He has spent 28 years at Illinois, where he is a Research Professor in Mechanical Science and Engineering and a Senior Technical Associate Director at NCSA, leading the Research Consulting directorate to address critical challenges in academia and industry.
Early in his career, Dr. Koric developed widely adopted numerical methods for multiphysics modeling of steel solidification in continuous casting. From 2014 to 2020, he led pioneering projects on the Blue Waters petascale supercomputer, demonstrating scalable engineering applications on peta- and emerging exascale platforms. More recently, his research focuses on AI-driven, data-driven, and physics-informed deep learning methods to accelerate simulations and solve complex computational problems. He has authored over 100 publications and received numerous awards in high-performance computing and engineering.
Online only; join via Zoom.