Rolling is the core step of steel production. It defines the major properties of final products of the steel industry like coils, wires or tubes which are delivered to a wide range of different industrial sectors. Because of the enormous influence of the rolling processes into the product quality, in modern rolling mills high degrees in process supervision and efficiency are reached. Current sensor equipment at rolling facilities in steel production delivers masses of data and information about the process, the product and its quality. And this data are supplied at high sample rates and high spatial resolutions. But there is a gap between the applied high-tech equipment and the exploitation of such measurements: besides their use for online control purposes such measurements are not exploited for online description of their impact into the product properties. To make use of the high amount of online available data new technologies have to be applied. And those technologies can be found in the areas of Data Analytics and Big Data architectures and methods. Big Data includes the aspects of volume, velocity, variety and veracity of data, which are not met in all cases by steel production data. But Big Data concepts considers inherently the request for a fast and online data processing. This in turn fulfils the needs of a supervision of the fast rolling processes in steel industry.

Therefore, the technological related objectives of the project are

• To toughen up the data storage systems at the industrial sites by enhancing the existing solutions and to integrate innovative tools from Big Data technology, because the existence of proper data sources is a basic prerequisite to achieve the technical objectives.
• To establish methodologies for data transmission to achieve highest possible data throughput rates, because this becomes necessary for the aimed online or near online processing.
• To develop and apply newest technologies of Data Analytics for analysis and modelling, because the existing data acquisition systems supply masses of data of various structures and contents, which have to be analysed and processed together.

To find industrially applicable solutions for these objectives, three industrial application cases were investigated with different target parameters to be modelled:

• at the hot-rolled wide strip production of SZFG,
• at the cold rolling processes of Marcegaglia Ravenna
• at the wire rod rolling plant of JSW