ArcelorMittal - automating the quality assurance of steel
The client
ArcelorMittal Ghent produces premium quality flat steel products. A process that is highly automated because of the dangerous, dusty, and warm working environment. The automation in question is the Online Statistical Process Control (OSPC), an end-to-end process monitoring web application. With OSPC, ArcelorMittal delivers its customers the ultimate certificate of excellence since all its products are manufactured by a chain of processes that are monitored around the clock.
The added value of our consultant
Stijn Bettens
In the summer of 2022, Stijn Bettens joined the industrial IT, automation and models group to work on the OSPC application. The OSPC functions as a limit switch, that raises alarms when signal values cross pre- defined limits. Stijn’s task was to make sure those limits were clearly defined, by implementing the statistical models, confidence intervals, control limits, process capabilities and alarm triggers.
To implement the models Stijn made use of a C# .NET framework, Angular and Cypress. Next, all formulas were validated using pen and paper, and custom Python scripts. Throughout development, Stijn listened to the users of OSPC and transformed their ideas and requests into intuitive features.
The result is an unbiased and efficient estimator that automatically learns a process signal’s natural variation and detects special causes of variations. On top of that, it can cope with thermodynamic measurements, weight calibrations, mean times to failures, product specification assessment and so on. Finally, it’s flexible in dealing with manual and IoT measurements. Whether it's a single measurement every month or thousands per day, immediate alarms or long-term trends.
Currently, Stijn is helping roll out the OSPC in multiple countries and acts as a strategic sparring partner for ArcelorMittal, ready to take on project management tasks when needed.