DE
ENProject i-CAST
Intelligent tools to accelerate Digitalization in Continuous Casting of Steel by coupling Machine Learning to Advanced Sensors and Digital Twins
Overall approach
i-CAST is a collaborative research project bringing together leading steel manufacturers, research institutes and technology suppliers with the goal of using advanced digitalisation technologies to reduce or eliminate defects in continuous casting products. To achieve this goal, the project leverages state-of-the-art intelligent sensors coupled with cutting-edge digital twins of the casting process for online identification and defect correction in the cast product.
i-CAST will enable more intelligent process control, improve product quality, reduce scrapped production, and strengthen the competitiveness and sustainability of the European steel industry.
Within i-CAST, K1-MET will work on defect classification and the impact analysis of tramp elements on casting, the AI-based identification of defect locations and castability predictions, as well as the evaluation of project impacts and support of dissemination activities to maximize knowledge transfer across the steel sector.
Objectives
The i-CAST project aims to:
- Integrate cutting-edge sensors such as line-scan pyrometers, fibre-optic temperature- and mold oscillation sensors as well as surface topography measurements in a comprehensive digital platform
- Develop digital twins and accelerated process simulations of the casting process to deepen the understanding of solidification and defect formation mechanisms
- Combine process data with findings from surface inspection and microstructure analysis to improve understanding of defect occurrence and categorize defects in a standard classification framework
- Apply advanced machine learning and AI technologies for real-time identification and correction of defects on cast products, improving product quality, process reliability, and production yield
