Cornell University

11/18/2026 | Press release | Distributed by Public on 11/18/2026 13:54

CHESS collaboration earns national award for advancing AI-driven experiments

A collaboration involving the Cornell High Energy Synchrotron Source (CHESS) has received a 2026 HPCwire Editors' Choice Award for connecting artificial intelligence, high-performance computing and scientific instruments across the country to enable autonomous experiments.

The team was recognized in the category of Best HPC/AI Collaboration (Academia/Government/Industry) for developing a nationally distributed experimental infrastructure through partnerships among CHESS, Oak Ridge National Laboratory (ORNL), the University of Tennessee, Knoxville, the University of Utah, the National Institute of Standards and Technology (NIST), the National Science Data Fabric (NSDF) and ViSOAR LLC.

The publication's annual HPCwire Readers' and Editors' Choice Awards recognize achievements in high-performance computing, artificial intelligence and data-intensive research.

The award-winning collaboration demonstrates an approach to synchrotron experiments in which AI systems analyze data and direct subsequent measurements in real time, even when scientific instruments and computing resources are located hundreds of miles apart.

"This collaboration is exciting in how it brings together expertise and capabilities from different research institutions," said Gabriella Carini, director of CHESS. "Combining X-ray measurements with neutron data, physics simulations and advanced computing opens new ways to investigate materials. CHESS provides an important experimental component of that larger research effort."

The team demonstrated the technology earlier this year at CHESS's Structural Materials Beamline, using high-energy X-rays to examine internal stresses in a metal sample. Data collected at CHESS in Ithaca were transmitted to ORNL/NSDF computational resources, where an AI-driven active-learning system analyzed the results and identified which measurements would provide the most valuable additional information. Instructions were then automatically sent back to CHESS, creating a continuous feedback loop between the experiment and remote computing infrastructure.

The demonstration combined technologies developed by the collaborating institutions, including ORNL's INTERSECT autonomous workflow platform, NIST-developed AI active-learning capabilities and NSDF-ViSOAR data infrastructure.

By connecting these capabilities, the collaboration extends CHESS's longstanding approach of bringing together researchers with different scientific backgrounds, including X-ray specialists and scientists studying specific materials and research questions.

"At CHESS, we bring together researchers with expertise in X-ray techniques and scientists who specialize in different fields," said Amlan Das, staff scientist at CHESS. "This collaboration takes that a step further by connecting expertise and resources across institutions. We aim to lower the barriers to conducting complex experiments and help researchers make better use of their time and data."

The project builds on years of work at CHESS developing beamline software, instrument controls and data-processing systems. Initially developed to support increasingly complex experiments and manage large datasets, these systems now provide the foundation for autonomous experimentation.

"Much of this work began because we needed better ways to manage instruments and data," said Werner Sun, IT director at CHESS. "Now we're connecting those systems with tools developed by our partners to automate processes that previously required manual intervention."

For Das and Sun, the award is also recognition that the work being done at CHESS has applications beyond the facility.

"It's particularly meaningful to be recognized by the scientific computing community, which is outside my primary field," Das said. "Computing infrastructure supports much of the science we do at CHESS, so it's encouraging to see that work recognized for its broader impact."

"The award was a big surprise," Sun said. "It's validating to know that the work we've been doing over the past several years is being recognized and could have an impact beyond CHESS."

The team plans to continue refining the infrastructure, integrating additional components and making autonomous experimentation more accessible to researchers. Eventually, Sun explains, users could bring their own AI models alongside their samples, connecting them to CHESS instruments and computing resources at other institutions.

Although the initial demonstration focused on structural materials, the same infrastructure could support research in other areas, including quantum materials and biological systems.

Additional CHESS team members involved in the collaboration include Chris Budrow, Sven Gustafson, Matt Miller, Diwakar Naragani, Kelly Nygren, Katherine Shanks, Keara Soloway, Rolf Verberg and Arthur Woll.

Cornell University published this content on November 18, 2026, and is solely responsible for the information contained herein. Distributed via Public Technologies (PUBT), unedited and unaltered, on November 18, 2026 at 19:54 UTC. If you believe the information included in the content is inaccurate or outdated and requires editing or removal, please contact us at [email protected]