Department of Materials AI & Big-Data
- Head of Division
Kang, Seong-hoon
Department Introduction
Department of Materials AI & Big-Data focuses on development of digital transformation technologies for innovative materials and processes. To this end, the research at Department of Materials AI & BigData covers artificial intelligence technology, virtual engineering technology, data generation/collection/ analysis/management system.
Major Activities
- Automatic data correction and analysis : 4 papers published and 2 software registered
- Deep learning-based artificial intelligence for super-resolving EBSD images: 1 paper published
- Reconstruction of a 3D microstructure geometry from 2D microstructure images: 1 software registered
Major Research Area
- Development of micro pattern classification technology based on convolution neural network with class activation map
- Investigation on physical/mechanical properties and chemical composition of new materials based on Bayesian optimization method.
- Development of technology for super-resolving microstructure image data
- Development of technology for automatic defect correction, automatic analysis, and automatic collection of research data
- Development of autonomous material design platform based on data and artificial intelligence
Future Research Plan
- Flash optimization and prediction in hot forging process based on deep learning
- Development of an artificial AI-based autonomous material design platform that can be used under small data conditions
- Development of automatic processing and analysis programs for structured and unstructured data
- Quality prediction of an ultrathin sheet in a cold rolling process using artificial intelligence
- Weld residual stress prediction using a deep learning architecture
Major R&D Activities
Development of data-based artificial intelligence technology for material and material process development.
- Develop a system for automatic correction, analysis, and collection of research data to build a high-reliability research DB, and develop artificial intelligence technology for effective material design and material process development using the data built through this
- Development of correction technology for optical microstructure error caused by non-uniform light source
- Development of correction technology for optical microstructure error due to non-uniform focus
- Development of automatic phase classification technology from multiphase optical images
Development of virtual engineering platform for vehicular lightweight materials
- A virtual engineering platform to minimize cost and maximize efficiency in developing materials and parts via microstructure-based multi-scale simulation
- Development of technology to predict the forming limit diagram of lightweight sheets and related softwares
- Development of technology to predict 3D casting microstructures and related softwares and modules
- Development of artificial intelligence-based technology to predict edge flangeability