About this Abstract |
| Meeting |
2027 TMS Annual Meeting & Exhibition
|
| Symposium
|
AI-Enabled Materials Processing: Integrating Accelerated Experimental Workflows and Processing-Aware Machine Learning
|
| Presentation Title |
From Automated Materials Experiments to Trusted Autonomous Discovery in AIMD-L |
| Author(s) |
Christopher Stiles, Rik Bose, Sam Scheck, Steven Storck, Ali K. Shargh, Piyush Wanchoo, Ashwini Gupta, Lori Graham-Brady, K. T. Ramesh, David Elbert, Todd Hufnagel, Jaafar El-Awady, Ali Rachidi, Kacper Kowalik, Harrison Park, Alexander deJong, Gannon Murray, Arjun Sreedhar, Konrad Muly, Lucas Rackers, Axel Krieger |
| On-Site Speaker (Planned) |
Christopher Stiles |
| Abstract Scope |
Agentic AI is rapidly expanding what is possible in automated materials laboratories, but the key challenge is building autonomy that is robust, adaptable, and scientifically grounded. The Artificial Intelligence for Materials Design Laboratory (AIMD-L) provides a platform for advancing from automated experimentation toward trusted autonomous materials discovery. AIMD-L integrates high-throughput processing, characterization, mechanical testing, data infrastructure, computational prediction, and machine learning within a predict-acquire-measure-learn cycle. Capabilities include MAXIMA for automated X-ray diffraction and fluorescence, SPHINX for high-throughput nanoindentation, and HELIX for laser-driven micro-flyer impact testing with photon Doppler velocimetry. Together, these tools generate multimodal datasets linking composition, microstructure, and mechanical response. This presentation discusses how AIMD-L is shifting autonomy metrics from throughput toward scientific understanding gained per action, accounting for uncertainty, cost, constraints, and value of information. The goal is co-science, where human expertise actively trains, guides, and improves autonomous discovery systems. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Machine Learning, Characterization, Extraction and Processing |