Modern automatic surface inspection systems (SIS) supply a very large amount of detailed information concerning the presence of defects on strip surfaces. The mass data thereby accumulated frequently remain unused with regard to selective optimisation of the SIS or process.
Comprehensive use of such strip data is necessary for the purpose of so-called coil grading, monitor-ing, checking the plausibility of the SIS findings, and tracking the specific strip quality, for example.
The BFI has created an efficient evaluation software tool (DataSIS) whereby the mass data of an SIS can be analysed simply and flexibly according to practice-related terms of reference, also over lengthy production periods.
The findings and data transparency obtained in this respect are an important prerequisite for subse-quent process optimisation.

Process data correlation

It is largely known that the causes of surface defects can lie at various distributed points of the process chain. For this reason it is necessary, as a first step, to collate the SIS data in relation to the strip length together with the process data from previous process stages. This places particular demands on the data systems engineering, because quite differing data from the process chain have to be ag-gregated for one section of the strip product. Support through special tools (e.g. DataTools) is indis-pensable in this regard. Once the data have been collated, they can be used to seek correlations be-tween SIS and process data by means of data mining methods.