About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Printed Electronics and Additive Manufacturing of Advanced Functional Materials and Devices—From Processing Concepts to Applications
|
| Presentation Title |
Optimizing Inkless Printed Electronics With Machine Learning Models |
| Author(s) |
Colton Bevel, Masoud Mahjouri-Samani, Sumner Harris |
| On-Site Speaker (Planned) |
Colton Bevel |
| Abstract Scope |
We present a machine learning approach to optimize inkless printed electronics so that users may enter a target resistance and receive optimized printing parameters to achieve it. The process is based on a dry additive nanomanufacturing (Dry-ANM) technique that produces pure nanoparticles via pulsed laser ablation. These particles are directed via carrier gas through a nozzle and sintered in-situ onto a substrate. For this study, we constructed a digital twin from printed-silver data and used it to benchmark active learning configurations that combined different surrogate models, acquisition functions, and batch sizes. We implement the selected model in a fully autonomous, closed loop process for material discovery and demonstrate its capability to print recycled copper. Ultimately, it enables rapid data-driven development of custom-printed circuits for electronic applications using recycled material as feedstock. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Additive Manufacturing, Machine Learning, Electronic Materials |