machine learning framework
examples
How can we make better steel?
Producing high quality steel is a must for voest alpine stahl GmbH to satisfy its customers. Continuously improving the quality ensures to be a leading produce of high quality steel. Steel production is a complex process which involves several production steps.
One of these is continuous casting where the liquid steel is casted in slabs. A problem which often occurs in continuous casting is the so called clogging which may causes impurities in the produced slabs.
How can one decrease the amount of clogging?
What are possible countermeasures?
To answer these questions a system for semi-automatic data analysis based on the machine learning framework for Mathematica has been set up at voest alpine stahl GmbH.
This so called "clogging analyzer" automates the analysis of data from over 50 different casting scenarios.
The results of these many modeling tasks are summarized in such a way that the steel making experts can derive possible countermeasures to decrease the clogging effect.