Interview: Artificial Intelligence and cLynx at SAUERESSIG
A look behind the scenes at SAUERESSIG: How does the work on AI projects in our company really look like? Our AI-expert Daniel Schmidt talks about his experience and fascinating projects with intelligent technology.
1. Mr. Schmidt, at SAUERESSIG you work on projects to make machines more intelligent. What exactly is it all about?
Through the digitalization and the increasing use of electronic devices, we all generate more and more data. But in the end we are often only interested in a few specific information. In some cases, we can access these using clear, predefined rules. In most cases, however, what we want to know is hidden in the data and can only be obtained through a complex set of rules. Recognizing and writing down this set of rules is very complicated, even for humans, and not always possible.
To solve the problem, the field of artificial intelligence emerged. This is a field of computer science that tries to solve these problems and whose results appear intelligent at the end. This field overlaps with a variety of other disciplines like mathematics, neuroscience and logic or particular in our case, the field of computer vision.
2. What is computer vision in that context?
Computer vision generally includes all processes that enables a machine to “see”. In effect, this means that information can be obtained from images. This includes both the classic algorithms in image processing and, in recent years, an increasing number of developments in artificial intelligence and, in particular, deep learning. As a result, a wide variety of visual tasks can be solved in the end, some of which appear simple to humans, but have a very complex set of rules. These include for example, autonomous driving, automatic writing and character recognition.
3. Where is artificial intelligence already being used at SAUERESSIG?
Artificial intelligence and especially computer vision are particularly relevant in the printing industry, because in most cases the final product is a printed image. This image is adjusted for each next step throughout the entire course of processing from design to printing.
Therefore, the aim of our first product was the quality control of the last step in the cylinder production. This has meanwhile developed into a whole family of programs that, under the name cLynx, supports quality control at various points in the company and, in cooperation with our colleagues, ensures consistently high quality for our customers.
4. Where does your passion for data streams and artificial intelligence come from?
During my computer science studies, I had the opportunity to get to know many sub-areas of this broad field. I quickly became fascinated by the possibilities of artificial intelligence and computer vision in particular. In addition to the technical understanding, imagination and creativity are also required in order to use the knowledge gathered to solve the specific problems. Due to the fast developments in artificial intelligence you constantly work on new challenges, for which there were often no solutions until then. That makes the project new and interesting, because you don’t always know if it will be solvable at all.
Through the collaboration with Roman Gevers and Professor Javier Villalba-Diez, I had also the opportunity to further explore the current developments in artificial intelligence in a doctoral thesis. In this context I was already able to publish some scientific papers.
5. That´s sounds versatile and exciting. What exactly are your scientific publications about?
Basically, the questions is how the current developments in artificial intelligence will affect the work of tomorrow in general and particular. It is important to consider both basic techniques as well as specific approaches to solving problems with the help of AI in order to get a comprehensive picture of what is currently possible and where developments are heading.
6. What new developments are you currently working on today?
We are currently in the development of some larger projects, which we will report on in the next few month. Our “Similarity Search” is currently in operation. With this software solution it´s possible for our customer service to find similar previous orders based on an image. In addition, we are in the test phase and further developing our “AI Screening”. We want to combine the advantages of automatic and manual screening and offer the customer an added value.
7. And what do you do when you are not working with data?
Even if some colleagues have the idea that I deal with data all day, I often get out into nature in my free time. Whether on foot or by bike. I also enjoy discovering new music and cooking. And as soon as traveling becomes easier, I like to explore new places again and get inspired by new ideas.
Learn more about cLynx:
Click here to access the papers:
1. Deep Learning for Industrial Computer Vision Quality Control in the Printing Industry 4.0
2. Characterization of Industry 4.0 Lean Management Problem-Solving Behavioral Patterns Using EEG Sensors and Deep Learning
3. Geometric Deep Lean Learning: Deep Learning in Industry 4.0 Cyber–Physical Complex Networks
4. Industry 4.0 Lean Shopfloor Management Characterization Using EEG Sensors and Deep Learning
5. Deep learning enabling quality improvement in rotogravure manufacturing