AI & Digital Technology in Steelmaking

The steel industry is among the largest and most data-rich manufacturing sectors in the world — a modern integrated steel plant generates terabytes of process data every day from thousands of sensors, thermocouples, flow meters, spectrometers, and level gauges. Until recently, most of this data was used for immediate process control and then discarded. Artificial intelligence — particularly machine learning (ML) and deep learning — is now being applied to extract insights from this historical data to optimise processes, predict quality, prevent failures, and reduce energy consumption.

The applications of AI in steelmaking span the entire value chain: from blast furnace burden optimisation and BOF endpoint prediction, through continuous caster quality prediction and rolling mill speed optimisation, to finished product quality grading and supply chain logistics. While the technology is genuinely transformative, steel companies have found that the practical challenges — data quality, process variability, model maintenance, and operator trust — are as significant as the technical ones.

This is not a distant prospect. POSCO, thyssenkrupp, SSAB, ArcelorMittal, Baowu, and virtually every major steelmaker now has dedicated AI/digitalisation programmes. Some are in production use; many are in extended pilot phases. The industry is in early-to-mid adoption, with significant heterogeneity between leading players and the majority.