Ever since his youth, Haug had wanted to find out how to get computers to do what he wanted. He hadn’t given a PhD a second thought, but Stefan Wagner managed to convince him through his lectures at the University of Stuttgart, where he was still teaching and conducting research at the time. His work as a PhD student led him to a software engineering conference in Crete, where he met his current project partner from Zeiss. On the way back from the event, they shared a taxi to save money: “I told him about my actual topic. He was facing similar challenges at the company, and that’s how we came up with our joint Software Campus project,” Haug recalls.
First the Applications, Then the Project
To begin with, applications and a detailed project description had to be written. Then came the approval – and with it, funding of 115,000 euros for the project team from the Federal Ministry of Research, Technology and Space. “The Software Campus has a unique structure – there, you can put ideas into practice together with an industry partner and bring science out of the ivory tower and into reality.” This gave the young scientist the chance to see whether his work was relevant in practice and reflected the problems people face. He also learnt important skills such as team leadership and self-organization, which will help him in his future career.
At Zeiss, the project team found more than one starting point for integrating AI components into processes. “Following interviews with staff, we identified 13 issues relating to production, data security and quality assurance,” explains Haug. The company, headquartered in Oberkochen, Baden-Württemberg, specializes in optics for semiconductor manufacturing and has been very successful in this field. The Dutch machine manufacturer ASML, the global market leader in semiconductor manufacturing equipment, uses only Zeiss optics.
In semiconductor manufacturing, the smoothness of the optical components presents a particular challenge. If one were to scale up one of the mirrors to the size of Germany, the highest elevations must be no more than a few millimeters; otherwise, the required precision cannot be achieved. Consequently, both the materials and the processes must meet the very highest standards. Automation using AI can save time and money here. This ensures efficient production and avoids unnecessary waiting times.
Everyone Benefits
All parties are more than satisfied with the results: “The collaboration was excellent. There was a great deal of interest in our project. A meeting was even held on-site at the company, attended by all project participants, who provided positive feedback.” In addition to optimizations in production processes, the issue of secure data transfer on cloud-based platforms also played a key role. All of this culminated in a scientific publication entitled “MLOps Adoption in the Manufacturing Industry: A Case Study with Zeiss SMT”.
Ultimately, it is a win-win situation for both sides: “Zeiss benefits directly from our results, and for me, the Software Campus network has opened up new doors,” says Haug. The paper also brings him closer to his main goal: “Once the Software Campus project concludes at the end of April, I will submit my dissertation shortly afterwards. After that, I can well imagine moving into the business world and helping other companies with the integration of AI.”