DigiTube - Digitized in-line measurement and analysis of surface defects on tubes.

Purpose and goal:

In the project, various techniques for one company specific case with surface defects on steel tubes, have been tested and the advantages and disadvantages of the various techniques have been studied and compared. Furthermore, AI models have been developed for identification of surface defects on steel pipes. In this work, we have investigated how AI models can be optimized, but also how the measurement system can be designed and improved to facilitate the classification of surface defects.

Expected results and effects:

The project partners have tested their different measurement techniques in an application case from Alleima. In this application case, we have investigated the possibilities of different imaging techniques in so-called fault detection, where we try to image and characterize defects, in this case over-rolling, on steel tubes. During the project, we have done a comparative study of different imaging techniques on the same tube sample, where the results have been analyzed and examined in the context of combining data sources to increase accuracy in measurement, when using data fusion.

Approach and implementation:

The project consortium and the participants from each project partner were very well suited to run this project. A great personal commitment has dominated the work with many active discussions and focus on solutions from the parties. However, the procedure of carrying out measurements on the same tube sample turned out to be considerably slower than planned for and with extra time for each participant it contributed to the fact that we could not follow the original time schedule and the project had to be extended with a few months.