2025 MRS Fall Meeting & Exhibit

Symposium MT01-Accelerating Materials Research Beyond Data-Driven Approaches—Physical Knowledge and Human Intervention in Autonomous Experiments

Material discoveries for improved societal, environmental, safety device applications etc. often require optimization of expensive experimental processes for materials synthesis, characterization, and learning structure-property relationships over large parameter and function spaces where the exhaustive grid or random search are too data intensive. This resulted in strong interest towards automated and autonomous exploration through machine learning (ML) guided experiments. Also in the practical setting, the material scientists either possess partial prior knowledge of the systems or gains knowledge with on-the-fly continual learning from the experiments. This can be leveraged in further increasing efficiency and alignment of autonomous experiments with combining known physical knowledge and minor intervention of domain experts into purely data-driven ML methods.
This symposium aims to explore the impact of Artificial Intelligence (AI) and ML in material science research, which covers the topics that includes design and development of novel ML approaches beyond purely data-driven approaches in material characterization and synthesis. This symposium particularly covers topics such as the limitation of purely data-driven approaches in autonomous experiments and encourages the strategic design and development of novel algorithms for physics knowledge injection (physics-AI) and domain expert intervention (human-AI). The symposium also covers development and application of natural language processing (NLP) methods in material research. The goal of this symposium is to bring together researchers across the material science and ML communities to foster and accelerate the development of novel techniques to not only advance the fundamental knowledge of material science domains but also to highlight the collaboration in algorithm development to enhance physics and human AI cooperation.

Topics will include:

  • Limitation of conventional data driven approaches in automated experiments.
  • Autonomous synthesis, characterization and modeling beyond purely data-driven methods
  • Improving AI alignment over human preferred automated experimentation.
  • Physics-Informed Machine Learning
  • Physics knowledge injection in data analysis and autonomous experiments.
  • Autonomous systems for materials research with human intervention.
  • Latent space learning in physical knowledge extraction and discovery
  • Decoding physically constrained latent space.
  • Natural Language Processing (NLP) in material research.
  • Materials informatics.

Invited Speakers:

  • Rigoberto Advincula (The University of Tennessee, USA)
  • Sterling Baird (University of Toronto, Canada)
  • Hannah Noa Barad (Bar-llan University, Israel)
  • Yasemin Basdogan (University of Rochester, USA)
  • Silvana Botti (Ruhr-Universität Bochum, Germany)
  • Alex Bourque (3M, USA)
  • Henry Chan (Argonne National Laboratory, USA)
  • Jean Yves Delannoy (Arkema, France)
  • Kedar Hippalgaonkar (NanyangTechnological University, Singapore)
  • Klavs F. Jensen (Massachusetts Institute of Technology, USA)
  • Rinkle Juneja (Oak Ridge National Laboratory, USA)
  • Boran Ma (The University of Southern Mississippi, USA)
  • Adrian Del Maestro (The University of Tennessee, Knoxville, USA)
  • Tyler Martin (National Institute of Standards and Technology, USA)
  • Marshall McDonnell (Oak Ridge National Laboratory, USA)
  • Fanyang Mo (Peking University, China)
  • Joydeep Munshi (GE Aerospace, USA)
  • Lena Simine (McGill University, Canada)
  • Shuwen Yue (Cornell University, USA)

Symposium Organizers

Arpan Biswas
The University of Tennessee, Knoxville
USA

Arthi Jayaraman
University of Delaware
USA

Yongtao Liu
Oak Ridge National Laboratory
USA

Jie Xu
Argonne National Laboratory
USA

Topics

artificial intelligence autonomous research informatics machine learning