Three New Publications from SEAD Lab

We developed a method for the design of high-dimensional materials by integrating physics-based, uncertainty-aware neural networks with Bayesian optimization. This method was applied to the design of kirigami metamaterials to achieve a near-zero Poisson’s ratio and support conformable wearable applications.

• Jinyang Li, Jyotshna Bali, Ko-Lin Wang, Suyi Li, and Jie Chen, High-Dimensional Design of Kirigami Patches Metamaterials Using Physics-based Machine Learning with Uncertainty Quantification. Advanced Engineering Informatics 2026.
  
• Jyotshna Bali, Jinyang Li, Jie Chen, and Suyi Li, Rapid Design and Fabrication of Body Conformable Surfaces with Kirigami Cutting and Machine Learning. Advanced Science 2026.
  
  

We developed a method that adapts models trained on simulation data to experimental data while estimating predictive uncertainty. We show the effectiveness in high-throughput materials characterization. The method improves accuracy, detects out-of-distribution samples, and supports trustworthy ML-driven materials discovery.

• Jie Chen, Timothy Long, Michael Wall, Todd Hufnagel, and Wei Chen. Uncertainty aware machine learning for bridging simulation and experiment in high throughput materials characterization. Scientific Reports 2026.
  
  Congratulations to all authors and collaborators!