New research project: 20 to 30 per cent more energy efficiency for the metal processing industry thanks to artificial intelligence

The metal processing industry requires a considerable amount of energy, especially for sawing, grinding and milling. Using artificial intelligence methods, researchers are now hoping to increase efficiency by 20 to 30 per cent. The SAELING project is realised jointly by the University of Klagenfurt, the KU Leuven and the industrial partners Voestalpine and Siemens and is funded by the Austrian Research Promotion Agency FFG at a cost of around 1.2 million euros.

Voestalpine uses around 2,500 sawing, grinding and milling machines in its industrial plants. These consume approximately 21 GWh per year, corresponding to the electricity consumption of around 4,750 average Austrian households.

“Metal processing machines on the factory floor fulfil a variety of tasks. At present, the question of which machine should be used for which task and when has yet to be definitively resolved,” states Gerhard Friedrich, head of the SAELING project at the Department of Artificial Intelligence and Cybersecurity at the University of Klagenfurt. “We need to take many factors into account in order to develop strategies for sawing, grinding and milling in these kinds of workshops in a way that saves energy and resources wherever possible. Considering and simulating these factors along with their full impact is beyond the capabilities of human reasoning. In particular, the behaviour of these machines cannot be described with sufficient precision, but rather it has to be learned for the purpose of optimisation.”

Artificial intelligence methods are now set to significantly reduce energy consumption thanks to more efficient use, as Gerhard Friedrich goes on to explain: “Approaches such as reasoning, optimisation and machine learning will be put to use.”

The researchers’ primary use case is Voestalpine, where they hope to achieve an energy efficiency increase of 20 to 30 per cent, i.e. around 4 to 6 GWh per year. The results from SAELING should facilitate analogue savings in other production areas. In addition to CO2 emissions, it should also be possible to reduce lubricant consumption, for example. It is intended that the tools developed in the project will be adaptable and can be extended to other areas of application, e.g. at SAELING’s partner Siemens.

The project name SAELING stands for SAving Energy by Learning and ImproviNG logic-based optimization models. SAELING was launched in early May 2024 and is funded by the Austrian Research Promotion Agency FFG to the tune of around 1.2 million euros. The total project costs amount to around 1.8 million euros.