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

Automated quality control methods will be tested and evaluated on a pilot scale. The project is based on the production of steel pipes at Sandvik and the surface defects that occur in the production. Based on the work in the case of tube manufacturing, a methodology for handling large amounts of measurement data from machine vision systems will be developed. The work includes evaluating and adapting methods, techniques, and measurement data quality for how data can and should be sorted, packaged, analysed, and adapted to a customer-specific user environment, in the process industry, using AI.

The results of the project will allow Sandvik to replace manual inspection of surface defects with automated quality control with machine vision systems in combination with an AI-based method for data management.

The system vendors participating in the project will be able to improve their products through increased knowledge of how data management for machine vision can bring increased customer value.

For Sandvik, the project results in the future may allow, with the help of data management and feedback to previous steps in the manufacturing process, to steer production towards fewer defects and thus reduce the need for post-processing and possibly reduced disposals.

The project is based on the production of steel pipes at Sandvik and the surface defects that occur in the production.

The goal is to develop a solution for automated pipe inspection which contributes to increased competitiveness through more efficient production as well as an opportunity to reduce disposals, complaints, and better use of raw materials.

System vendors are strengthened by being able to expand their customer offerings through knowledge of how data management for machine vision can bring increased customer value. Automated quality control methods will be tested and evaluated in the pilotline.