Self-X Artificial Intelligence for Steel Use Case

Exciting news for the steel industry! The integration of the s-X-AIPI concepts into the Electric Arc Furnace (EAF) is set to improve decision making through the use of self-adapting AI solutions. The properties of charge materials for EAF can vary over time, and it is essential to fulfil the steel grade-specific target analysis. With the help of machine learning algorithms like autoencoders, deviations in predicted and actual liquid steel status can be detected and corresponding changes can be applied to the scrap property parameters, data pipelines and/or to the modelling and optimisation tools. Additionally, autonomous AI models will be trained to predict critical elements in liquid steel and their confidence interval. Furthermore, a multivariate outlier detection algorithm will validate the input and output of the modelling and optimisation tools to avoid false decisions. With these advancements, the steel industry can achieve better control over the EAF operation and optimize the energy and resource consumption, particularly when using low-quality scrap types. The future is looking bright for the steel industry!