Kapernikov enables Umicore to increase plant capacity
Kapernikov helped Umicore to make its precious metal recycling processes more efficient. This enable ...
What happens if you put dozens of brilliant minds together in one place for 48 hours to think about the future of the smart factory? No, this is not the start of a bad joke. It’s the starting point of The Challenge, ArcelorMittal Belgium’s very first hackathon, taking place from September 15 to 17 in Ghent, Belgium. Kapernikov is proud to take part in this event and the company is totally ready to explore new possibilities in the field of artificial intelligence for the Industry 4.0.
“This hackathon is a great way to challenge ourselves as a team,” says Rein Lemmens, partner at Kapernikov. “Data management and artificial intelligence in industrial settings are fields of expertise that fit our company like a glove, so we are looking forward to put our skills to the test this weekend.”
Steel industry
The first hackathon by ArcelorMittal Belgium challenges participants to come up with a creative solution to a steel industry-related problem. This time, the problem definition has to do with a cold rolling mill, a high-precision mill that is used to reduce the thickness of steel to the customer’s specifications.
In order to create a continuous cold rolling process, the steel coils are welded together at the entrance of the mill, thus creating a big chain to continuously feed the rolling mill. In case the weld would break, the entire process would come to a stop, material would be lost, and there would be a high risk to damage the machine as well.
The Kapernikov challenge
Can you predict if a weld will break or not? And if so, can you highlight those areas that pose a risk? This is the challenge that Kapernikov is eager to take on during the upcoming hackathon. Kapernikov will be able to use an extensive data set of height images generated by the rolling mill’s 3D laser system to analyze the weld profiles.
When you are constrained by time, it’s critical to be well prepared. This is why in the past few weeks, the Kapernikov team has been gearing up well for the upcoming weekend. “We have been able to do a number of important tests on a small data set, which allows us to gain more insight into the problem,” says Maarten De Munck, Industrial Automation Engineer at Kapernikov. “We have also been working on ways to convert the data into a more convenient format, with the intention to work as fast and efficient as possible during the hackathon weekend.”
Learning and fun
While the Kapernikov team is plotting its way to glory, the company mainly looks at the hackathon as a fun learning experience, says Rein Lemmens. “We have the opportunity to dig deep into a specific problem for 48 hours and use the techniques that are really at the heart of our company. But next to a great learning experience, we also see this hackathon weekend as a way to get together and have fun with the people in our machine learning community.”