Symposium S.CT01-Artificial Intelligence for Material Design, Processing and Characterizations

The life cycle of materials processing involves several steps summarized as design, synthesis, testing/characterization, and application or usage. Traditionally, each step of this life cycle is conducted by humans, including analysis/interpretation of data that is generated during the simulation, experiment, and application as well as decision making. These data often span different length scales of materials from molecular and macroscopic scales and various time scales. They are usually high-dimensional, large in number, and heterogeneous in nature. All of their features make them difficult for researchers to obtain comprehensive insight into these data. Thus, this strictly human-centric material discovery and manufacturing process is clearly neither efficient nor guarantees optimal results. Recently, emerging data-driven techniques based on statistics, machine learning, and artificial intelligence (AI) have shown great potential for improving this process. They can unlock predictions of structures and/or performance of new materials with computations, assist in analyzing and interpreting (potentially extremely large) datasets generated in experiments, inform robots to perform autonomous research from synthesis/processing to structure-property characterizations, and yield operating instructions for manufacturing systems. To widen their application and speed up material and manufacturing innovation, this symposium will explore a broad spectrum of topics in which machine learning, AI, and robotics, which are of increasing importance in addressing challenges in material design, characterization and processing throughout the whole life cycle of materials. The interaction of humans and intelligent systems is a coordinated human-machine partnership. It will bring together researchers from interdisciplinary knowledge domains (computational/experimental materials, engineering, computer science, statistics, and robotics) to discuss the fundamental challenges in applying AI for materials and manufacturing in a close-loop manner as well as its future applications and impact. The broad implications resulting from the fruitful discussions will inspire researchers working across research fields to move forward and promote the basic knowledge development and technology deployment.

Topics will include:

Materials design assisted by machine learning
Machine learning augmented material characterizations
High performance computing methodologies for data-driven materials development
Data mining and machine learning from material literature
Interpretable machine learning for understanding physics-oriented models
Machine learning and statistics for material life cycle analysis (failure, recycle)
AI powered robot for autonomous experimental planning and research
Human-robot, human-intelligence system interactions in material processing and manufacturing

Invited Speakers:

Keith Brown (Boston University, USA)
Keith Butler (Science and Technology Facilities Council, United Kingdom)
Wei Chen (Northwestern University, USA)
Peter W. Chung (University of Maryland, USA)
Jacqui Cole (Cambridge University, United Kingdom)
Andy Cooper (University of Liverpool, United Kingdom)
Lee Crorin (University of Glasgow, United Kingdom)
Stefano Curtarolo (Duke University, USA)
Kurt De Grave (Flanders Make, Belgium)
Gerald Friedland (University of California, Berkeley, USA)
Matthew Hirn (Michigan State University, USA)
Sergei Kalinin (Oak Ridge National Laboratory, USA)
Benji Maruyama (Air Force Research Laboratory, USA)
Satoru Masubuchi (The University of Tokyo, Japan)
Shyue Ping Ong (University of California, San Diego, USA)
Rampi Ramprasad (Georgia Institute of Technology, USA)
Kristofer Reyes (University at Buffalo, The State University of New York, USA)
Patrick Riley (Google Accelerated Science, USA)
Joshua Schrier (Haverford University, USA)
Ichiro Takeuchi (University of Maryland, USA)