[IS] IS1 Frontiers in Structural Materials Research through Advanced Measurement Techniques

1:30 PM - 2:00 PM JST(4:30 AM - 5:00 AM UTC)

[IS7] Uncertainty-Aware AI-Based Indentation-to-Tensile Property Conversion for Local Mechanical Characterization

○Jong-hyoung Kim¹, Si Hyun Park¹, Deasik Kim¹, Im-Deok Kim², Seung-Kyun Kang², Ryuta Kasada³, Sanghoon Noh¹ (1. Pukyong Natl. Univ., 2. Seoul Natl. Univ., 3. Tohoku Univ.)

Keywords:

Indentation,Tensile property,Artificial neural network,Measurement uncertainty,Electron-beam welding

An uncertainty-aware AI framework for indentation-to-tensile property conversion is proposed. The framework incorporates experimentally quantified uncertainty through data augmentation, enabling noise-robust prediction with confidence estimates. It is validated using electron-beam welded ARAA steel.