Agenda

All times are in EDT.

TimeDetails

Tuesday, April 26, 2022

10:30 - 10:45 a.m.Welcome Remarks: Doug Kothe, Oak Ridge National Laboratory

10:45 - 11:45 a.m.Panel Discussion: Digital Twins and Artificial Intelligence - Overview of Needs and Challenges
Yousu Chen, Pacific Northwest National Laboratory
Scott Collis, Argonne National Laboratory
Matt Juckes, Aimsun, Inc.
Tom Kurfess, Georgia Institute of Technology
Matthew Nielson, GE Global Research
11:45 - 12:00 p.m.Group Photo
12:00 - 1:00 p.m.Working Lunch: Invited Speaker 1-1 - Saiph Savage, Northeastern University
The Future of A.I. for Social Good
1:00 - 2:00 p.m.Keynote Speaker: Vipin Kumar, University of Minnesota
Knowledge-Guided Machine Learning: A New Framework for Accelerating Scientific Discovery
2:00 - 2:15 p.m.Speaker 1-1: Farinaz Koushanfar, University of California San Diego
Holistic Co-Design and Optimization of Robust AI
2:15 - 2:30 p.m.Speaker 1-2: Guannan Zhang, Oak Ridge National Laboratory
Level Set Learning with Pseudo-Reversible Neural Networks for Dimension Reduction in Building High-Dimensional Twins
2:30 - 2:45 p.m.Speaker 1-3: Ravi Patel, Sandia National Laboratories
Error-in-Variables Modeling for Operator Learning
2:45 - 3:15 p.m.Break
3:15 - 4:15 p.m.Flash Talks
  • Samrat Chatterjee, Pacific Northwest National Laboratory: Constrained Reinforcement Learning for Safe Cyber Defense Optimization
  • Yigit Yucesan, Oak Ridge National Laboratory: Digital Twin Applications using Hybrid Physics-informed Machine Learning
  • Michael Churchill, Princeton Plasma Physics Laboratory: Simulation-based Inference Methods for Robust Digital Twins
  • Sharlotte Kramer, Sandia National Laboratories: Novel Physics-informed Neural Network Approach for Material Model Calibration
  • Pinaki Pal, Argonne National Laboratory: Neural Ordinary Differential Equations Approach for Time-series Prediction of Chemical Kinetics
  • Kishansingh Rajput, Jefferson Lab: Robust Digital Twin for Risk Averse Controller
  • Craig Bakker, Pacific Northwest National Laboratory: Domain Knowledge Incorporation and the Koopman Operator
  • Jaideep Ray, Sandia National Laboratories: Interpreting Data-driven Turbulence Closures using GLMMs
  • Massimiliano Lupo Pasini, Oak Ridge National Laboratory: Uncertainty-aware Predictions of Material Properties Using Graph Convolutional Neural Networks
  • Mariana Fazio, University of New Mexico: Autonomous Anomaly Detection in MeV Ultrafast Electron Diffraction
4:15 - 5:30 p.m.Breakout Sessions
5:30 - 7:30 p.m.Reception for onsite attendees

Wednesday, April 27, 2022

10:30 - 10:45 a.m.Day 2 Welcome and Introduction: David Womble, Oak Ridge National Laboratory
10:45 - 11:15 a.m.Invited Speaker 2-2: Christopher Rackauckas, Massachusetts Institute of Technology
The Continuing Advances of Differentiable Simulation
11:15 - 11:45 a.m.Invited Speaker: Stephan Mandt, University of California, Irvine
Deep Latent Variable Models for Sequential Data
11:45 - 12:00 p.m.Speaker 2-4: Nathaniel Trask, Sandia National Laboratories
Physics-Informed Multimodal ML for High-Throughput Scientific Discovery
12:00 - 12:15 p.m.Speaker 2-5: Mehmet Belviranli, Colorado School of Mines
Energy-Aware Execution of Neural Network Interference on Multi-Accelerator Heterogeneous SoCs
12:15 - 1:15 p.m.Working Lunch: David Stracuzzi, Sandia National Laboratories
Trusted Artificial Intelligence Research Campaign
1:15 - 1:45 p.m.Invited Speaker 2-4: Paris Perdikaris, University of Pennsylvania
Data-efficient Operator Learning with Quantified Uncertainty
1:45 - 2:15 p.m.Invited Speaker 2-5: Carianne Martinez, Sandia National Laboratories
How to Train Your Digital Twin: Practical Deep Learning Approaches to Modeling As-built Components
2:15 - 2:30 p.m.Speaker 2-6: Satish Karra, Los Alamos National Laboratory
Constraining Machine Learning with Physics-Based Codes
2:30 - 2:45 p.m.Speaker 2-7: Joseph Bakarji, University of Washington
Constraining Machine Learning Algorithms with Fundamental Theorems to Discover Differential Equations from Data
2:45 - 3:15 p.m.Break
3:15 - 4:15 p.m.Flash Talks
  • Gabe Fierro, Colorado School of Mines: Managing Semantic Linked-data Models for Enabling Digital Twins
  • Dong Chen, Colorado School of Mines: Safeguarding Data Privacy in Edge Computing-assisted Internet of Things
  • Balwinder Singh, Pacific Northwest National Laboratory: An End-to-End Workflow for Implementing Machine Learning Emulators in Energy Exascale Earth System Model (E3SM)
  • Isabel Scherl, University of Washington: Experimental Active Learning in the Loop
  • Aaron Young, Oak Ridge National Laboratory: Adrastea: A Framework for Efficient FPGA Design for Digital Twin Deployment
  • Ethan King, Pacific Northwest National Laboratory: Coupling Physics and Machine Learning to Model Temperature of Shear Assisted Processing and Extrusion
  • Majdi Radaideh, Oak Ridge National Laboratory: Time Series Anomaly Detection with Recurrent Neural Network Autoencoders for the Spallation Neutron Source
  • Amanda Howard, Pacific Northwest National Laboratory: Physics-informed Co-Kriging Model of a Redox Flow Battery
  • Dali Wang, Oak Ridge National Laboratory: Methods to Predict and Determine the Performance and Properties in Resistant Spot Welding
  • Urban Fasel, University of Washington: DesignOptimizaton and Data-driven Control of Flexible Structures Applied to Renewable Energy Systems 
4:15 - 5:30 p.m.Breakout Sessions

Thursday, April 28, 2022

10:30 - 10:45 a.m.Day 3 Welcome and Introduction: Pradeep Ramuhalli, Oak Ridge National Laboratory
10:45 - 11:15 a.m.Invited Speaker 3-6: Olga Fink, École Polytechnique Fédérale de Lausanne
Hybrid Operational Digital Twins for Complex Systems: Fusing Physics-Based and Deep Learning Algorithms for Fault Diagnostics and Prognostics
11:15 - 11:45 a.m.Invited Speaker 3-7: Mihai Anitescu, Argonne National Laboratory
Scalable Physics-based Maximum Likelihood Estimation using Hierarchical Matrices
11:45 - 12:00 p.m.Speaker 3-8: Kris Villez, Oak Ridge National Laboratory
Shape-Constrained Function Fitting as a Way to Satisfy Prior Knowledge
12:00 - 12:15 p.m.Speaker 3-9: Ben Hodges, University of Texas at Austin
Challenges and Opportunities for AI with Physics-Based Models of River Networks and Urban Flooding
12:15 - 1:15 p.m.Working Lunch: Ben Mintz, Oak Ridge National Laboratory
An Interconnected Science Ecosystem for Smart Laboratories of the Future
1:15 - 1:30 p.m.Speaker 3-10: Salvador Sosa Guitron, University of New Mexico
AI-Assisted Design and Virtual Diagnostic for the Initial Conditions and Operation of a Storage-Ring-Based Quantum Information System
1:30 - 1:45 p.m.Speaker 3-11: Qizhi He, University of Minnesota
Enhanced Physics-Constrained Deep Neural Networks for the Redox Flow Battery Modeling
1:45 - 2:00 p.m.Speaker 3-12: Po-Lun Ma, Pacific Northwest National Laboratory
Better and Faster AI-Assisted Aerosol-Cloud Processes in Earth System Models
2:00 - 3:00 p.m.Breakout Sessions
3:00 - 3:30 p.m.Break
3:30 - 5:00 p.m.Breakout Session Out-briefs
5:00 p.m.Adjourn