From vehicle parts and circuit boards to components for wind turbines - in industrial production, cameras and computers check the manufacture of products or components for defects on a daily basis. Particularly in safety-critical fields (of) the automotive industry or medical technology, where faulty brakes or implants can have life-threatening consequences, maximum reliability is required. The images captured by the cameras are increasingly being processed and evaluated using artificial intelligence methods. The automation of these processes is still in its infancy. Professor Dr. Tim Dahmen from Aalen University of Applied Sciences is now tackling this problem in the recently launched "EDI-Vision" research project. It is being funded by the EXPLOR program of the Kessler + Co. foundation for education and culture in Abtsgmünd.
"Wonder box" 386 DX
His first computer, a 386 DX with an incredible four megabytes of RAM, was the gateway to a new world for Professor Dr. Tim Dahmen when he was twelve years old. Day and night, he sat in front of the "wonder box" and taught himself how to program. And because the World Wide Web was still a thing of the future at the end of the 1980s, he regularly cycled to the neighboring village to get hold of the latest computer books in the local library. "The nerds in our school's computer club, who were three or four years older, were also a good source of information and you could learn a lot from them," recalls Dahmen. At 16, he entrepreneured his own software start-up and contributed central software components for health insurance billing to a cab company that specialized in ambulance rides. "That meant the shoeboxes with all the cab receipts were a thing of the past," says Dahmen, who has held the professorship for "Computer Vision and Data Science" in the degree programme in Computer Science at Aalen University of Applied Sciences since 2024. How did his parents react to the fact that he skipped school for the CeBIT computer trade fair and preferred employment with bits and bytes to everything else? "Ambivalent. They have already insisted that the boy needs to get out into the fresh air," smiles the Information scientist, who comes from near Saarbrücken.
Studies as an eye-opener
Even though studying medicine was on the cards after his Abitur, university entrance qualification, all roads led to computer science. This was at Saarland University, which is particularly known for its cutting-edge research in computer science and for the highly-respected German Research Centre for Artificial Intelligence (DFKI). "Studies were quite an eye-opener and a lot of fun. There were now solutions and explanations for many problems that had previously given me a bloody nose, so to speak," says Dahmen enthusiastically. After completing his German university degree, the now 47-year-old initially went into industry and worked as an advisor for SAP, among others. "That was a very travel-intensive phase in which I practically lived out of a suitcase, but it was also a very cool time. You get insights into many industries and projects that you might otherwise see in decades."
Interdisciplinary cooperation, cooperative venture, partnership, collaboration
During this time, he also came into contact with many researchers. "I was toying with the idea of doing a doctorate after all, and when I bumped into my old professor at CeBIT, it quickly became clear. A few weeks later, I started my doctorate," says Dahmen and laughs. For eleven years, the computer graphics specialist worked as a Head /Director researcher at DFKI, and the next logical step was the professorship in Aalen. The interaction with the "incredibly good students", who are self-organized and motivated, the discussions, the explanations - he enjoys it all very much. The fact that he can combine teaching with applied research here is a great benefit. "I'm interested in getting systems up and running, the application aspect is extremely important to me. And here in Aalen, we have a great research environment with very good industry contacts. The interdisciplinary cooperation, cooperative venture, partnership, collaboration with the many colleagues who are strong in research is also a great enrichment."
New image processing solutions
Dahmen has also come a step closer to his goal of linking research more closely with practice; practical experience. This is because he was recently awarded 50,000 euros from the EXPLOR funding programme, scholarship programme of the Kessler + Co. foundation for education and culture in Abtsgmünd, which he can use to advance his research into image processing solutions. "With the help of AI, great progress has been made in error detection in recent years, but data to train AI systems in industry is often very scarce. In addition, the decisions made by neural networks are often difficult to understand," explains Dahmen. "Such AI methods reach their limits when it comes to quality control or certification of safety-critical products." Although conventional digital) image processing has its advantages in this respect, programming it accordingly is extremely time-consuming.
"It's great when you can make a difference with your research"
"In my project, I want to combine the strengths of both approaches by using modern machine learning methods to automate the development of classic image processing programs that are easy to explain," says Dahmen. The passionate Information scientist is delighted that he is now being supported by the EXPLOR funding program. "It's great when you can make a difference with your research and the results benefit industry in the region in a very practical and application-oriented way."
