About this Abstract |
| Meeting |
2026 TMS Annual Meeting & Exhibition
|
| Symposium
|
Novel Strategies for Rapid Acquisition and Processing of Large Datasets From Advanced Characterization Techniques
|
| Presentation Title |
New Strategy of Surface Defect Detection in Metallic Coatings Using a Machine Learning-Based Model |
| Author(s) |
MinGyun Kim, Kisu Kim, Seunghee Roh, Donghwan Roh, Hwasung Yeom |
| On-Site Speaker (Planned) |
MinGyun Kim |
| Abstract Scope |
Metallic coatings play a vital role in enhancing the overall performance of base materials by improving their durability, mechanical stability, and resistance to environmental degradation such as oxidation and corrosion. Nickel-plated steel is an example which is commonly used in secondary batteries to improve the corrosion and oxidation resistance of base steel, offering superior durability. Thus, thickness and uniformity of the Ni coating are essential quality indicators but have been traditionally evaluated through non-destructive methods. However, these techniques involve high equipment costs, long processing times, and spatial constraints, resulting in trade-offs between reliability and productivity in industries. This study proposes a machine learning-based approach using hyperspectral imaging to detect surface defects. We aim at improving conventional detection accuracy while dramatically reducing inspection time. This ultimately offers a cost-effective and rapid quality assurance options to quality assessment industries of thin coatings. |
| Proceedings Inclusion? |
Planned: |
| Keywords |
Copper / Nickel / Cobalt, Machine Learning, Surface Modification and Coatings |