The Quality Inspector: Daniel Herfert develops industrial applications with AI inside

04. March 2019

The Quality Inspector

Daniel Herfert develops industrial applications with AI inside

Daniel Herfert, GFaI © WISTA Management GmbH

Daniel Herfert privately with his twins: Until he builds with robot toys, he has to be patient for a while. Image: © WISTA Management GmbH

He studied under a visionary. In 2050, a robot football team will defeat the German national team, Daniel Herfert’s academic teacher Hans-Dieter Burkhard loved to prophesise. Outlooks like this are not uncommon on the Adlershof campus of the Humboldt-Universität zu Berlin, and they fascinated the budding computer scientist from Schöneiche in Brandenburg. Herfert, now 38, would already spend his free time programming as a student. Everything mathematical and scientific captured his interest, and his branch of study was preordained.

The fact that he studied most of his major subjects at Burkhard’s Chair for Artificial Intelligence and then kept working there for another three years as a student employee, was all thanks to a pivotal experience on one “Long Night of the Sciences” in Adlershof around one and a half years earlier. Herfert saw quadruped robots playing “dog football” and was impressed. “Tangible research”: that is what makes robotics appeal to him.

Herfert has since remained loyal to the location Adlershof. He now heads the “Structural Dynamics and Pattern Recognition” department of the Society for the Advancement of Applied Computer Science (GFaI). Ten employees there are developing concepts aimed at supporting human quality testers in industrial production processes with artificial intelligence.

A key technology of the method is capturing and analysing sound data. Every automated production process generates sounds that differ minutely depending on whether a workpiece will come out a success or not. This is where the software “Wavelmage” comes in; developed and produced in Herfert’s department, it is capable of discerning differences in sound patterns. This way, from even the minutest deviations in the sounds of a machine, it can immediately identify faulty workpieces and automatically reject them.

The method is highly versatile and can be used, for example, in maintenance, quality assurance or non-destructive testing. In medical engineering, it helps detect malfunctions of heart support systems before it can come to fatal consequences. It also allows relatively precise predictions to be made over the foreseeable life span of assemblies, machines, plants and components. Working from the operating sounds alone, tool wear and malfunctions can be diagnosed before a human technician could possibly have noticed anything.

Herfert has worked at GFaI since 2010. It was his predecessor, the head of department he knew as a lecturer from his computer science studies, who approached him. Tennis is another passion that has been with him all his life. Most recently, he indulges in this sport on a court not far from his office, of the Berliner Tennis Club Wista in Adlershof. But isn’t his profession also playful in a lot of ways? “For sure,” says Herfert, a father of one-and-a-half-year-old twins. “I really look forward to when the kids are ready for the Lego ‘Robotics’ program.

By Winfried Dolderer for Adlershof Journal

Comments

Comments (0)

No comments found!

Your comment

Your comment is important to us

Related News

Adlershof Journal März/April 2019. Titelbild
Adlershof Journal March/April 2019
K. O. or A. I.? What artificial intelligence is doing to us and our time