Symposium MT02-Machine Learning in Action—Automated and Autonomous Experiments

The conventional materials innovation cycle heavily relies on human decision-making and manual operation of scientific tools, leading to slow progress. Pressing challenges like the electrification of everything, large-scale materials synthesis, waste stream upconversion, and energy conversion and storage demand a transformative approach to accelerate material discoveries. In this symposium, we aim to explore innovative methods that combine experimental automation and machine learning to conduct materials research at or beyond the state of the art. This convergence presents a unique opportunity for machine learning-driven autonomous experimentation, promising improved efficiency, accuracy, and reproducibility in materials synthesis and characterization, thus accelerating breakthroughs in materials and physics.

The symposium's primary focus is on showcasing the applications of machine learning in experimental tasks, with an emphasis on materials synthesis and characterization. The topics to be covered include automated and autonomous experiment workflow design, development of task-specific algorithms for experimentation, high-throughput synthesis and characterization, and the creation of digital twins for laboratories. By bringing together researchers from both the material science and machine learning communities, we aim to facilitate knowledge exchange, share recent advancements, and discuss the opportunities and challenges in this rapidly evolving field.

Topics will include:

Computer-vision based automated experiments
Modular high-throughput experiments
AI-driven autonomous experiments
Multi-fidelity workflow design
Algorithms for microscopy, spectroscopy, diffraction, and electrochemical experiments
Data-driven experiment planning, realization, and review
Automation beyond the benchtop, integration across the lab and countries
Digital twins and Ontologies in academic research contexts
Orchestration of autonomous campaigns with multiple tenants
Autonomous research data management

Invited Speakers:

Alan Aspuru-Guzik (University of Toronto, Canada)
Nicola Marzari (École Polytechnique Fédérale de Lausanne, Switzerland)
John Gregoire (California Institute of Technology, USA)
Milad Abolhasani (North Carolina State University, USA)
Benji Maruyama (Air Force Research Laboratory, USA)
Pinshane Huang (University of Illinois at Urbana-Champaign, USA)
Keith Brown (Boston University, USA)
Mahshid Ahmadi (The University of Tennessee, Knoxville, USA)
Steven R. Spurgeon (Pacific Northwest National Laboratory, USA)
Sebastian Siol (Empa–Swiss Federal Laboratories for Materials Science and Technology, Switzerland)
Jason Hattrick-Simpers (University of Toronto, Canada)
Alfred Ludwig (Ruhr-Universität Bochum, Germany)
Hannah-Noa Barad (Bar-Ilan University, Israel)
Kedar Hippalgaonkar (Nanyang Technological University, Singapore)
Yousung Jung (Seoul National University, Republic of Korea)
Eun-Ah Kim (Cornell University, USA)
Thomas Morris (Brookhaven National Laboratory, USA)
Dan Olds (Brookhaven National Laboratory, USA)
Kishna Rajan (University at Buffalo, The State University of New York, USA)
Esther Tsai (Brookhaven National Laboratory, USA)
Daniela Ushizima (Lawrence Berkeley National Laboratory, USA)
Rama K. Vasudevan (Oak Ridge National Laboratory, USA)
Yan Zeng (Lawrence Berkeley National Laboratory, USA)

Symposium Organizers

Yongtao Liu

Oak Ridge National Laboratory

Center for Nanophase Materials Sciences

USA

[email protected]

Andi Barbour

Brookhaven National Laboratory

USA

[email protected]

Lewys Jones

Trinity College Dublin

Ireland

[email protected]

Helge Stein

Karlsruhe Institute of Technology

Germany

[email protected]

Topics

autonomous research combinatorial electron microprobe scanning probe microscopy (SPM) scanning transmission electron microscopy (STEM) scanning tunneling microscopy (STM)