voestalpine wants AI to run 2,000 saws and mills on less energy

Austria’s voestalpine is using AI across production, from defect detection to energy use. Its largest stated target sits in the High Performance Metals Division (voestalpine).

SAELING: 2,000 machines, one energy target

The division runs around 2,000 saws, milling machines and grinders, using about 21 GWh a year, which voestalpine compares to the annual consumption of 4,750 Austrian households. The SAELING research project, run by voestalpine HPM DIGITAL SOLUTIONS in Kapfenberg, aims to raise their energy efficiency by 20 to 30% at the sites concerned.

SAELING started in May 2024 with support from the Austrian Research Promotion Agency FFG. Partners are the University of Klagenfurt, KU Leuven in Belgium, and Siemens. It builds on HPMconnect, a programme that already collects usage data from more than 250 connected machines.

Predicting cracks before they form

At voestalpine Stahl Donawitz, the company says machine learning is used “to predict longitudinal cracks in the billets before they have even appeared.” At voestalpine Tubulars, computer vision checks hollow billets for cracks and surface defects.

“AI provides us with important assistance in decision-making processes; the final decisions are made by our employees.”IT Manager, Metal Engineering Division, voestalpine

On the floor

Energy is where maintenance AI pays fastest in European plants: the saving shows up on the bill, not just in avoided downtime. SAELING is worth watching for one number: whether the 20 to 30% target holds outside the pilot sites.