At present, although entrepreneurs are aware of the potential of artificial intelligence to increase their competitiveness, many are reluctant to take a step, doubting the reliability of AI solutions. The Fraunhofer Institute for Mechatronic Systems Design (IEM) seeks to build trust in artificial intelligence systems. His teams have combined AI and control technology to create a completely reliable system: RoboGrinder, a grinder that eliminates up to 40% of grinding processes.
Founded in Munich on March 26, 1949, Fraunhofer-Gesellschaft is represented by industry and academia, the Bavarian government and the Federal Republic of Germany. engineers and scientists. A pioneer of innovative products, he holds numerous patents. 70% of his income comes from contracts with industrialists or specific contracts with the government.
Fraunhofer Institute for Mechatronic Systems Design
Mechatronics is a synergistic and systematic combination of mechanics, electronics, automation and real-time computing. It is used in all major industries: transport, consumer goods, capital goods… Robots, autonomous vehicles, ABS, hard drives, smart washing machines… are examples of system mechatronics.
Researchers at Fraunhofer IEM have set themselves the goal of counteracting skepticism and building trust in artificial intelligence, in particular through technologies that control these systems.
Stephen Coppert, head of the machine intelligence group at Fraunhofer IEM, explains:
“We combine proven control technologies with new artificial intelligence. While many control technologies are based on physical and mathematical models that can be analyzed and trusted, artificial intelligence, which in itself is not particularly reliable at first, usually relies solely on data and solves creative problems. »
RoboGrinder: intelligent grinder developed by the IEM Fraunhofer Institute
RoboGrinder is a concrete example of the reliability of an automated system that integrates AI. The institute developed it for the machine-building company Düspohl.
Until then, rolls for wrapping the profile, which have a very complex shape and often have to be ground, were ground there by hand, because automation using only control technology for rubber material is not easy to achieve. In fact, the material of the roller is so elastic that it bounces off after grinding, such behavior is very difficult to model and integrate into the approach to control technology. Although one might think that increasing the grinding process would be easier, the process, on the contrary, becomes inefficient. In addition, the surface quality of the rolls may deteriorate with the number of grinding processes. The use of AI solves these problems.
Stephen Coppert explains:
“We document all the grinding errors that occur during each grinding process and use it to teach the artificial intelligence model, so AI takes over the correction, especially during fine-tuning. In comparison, control technology acts as a barrier, and artificial intelligence ensures that the curve of the road is the best. »
In this way, the process can be reliably automated, which is the first. Companies engaged in the manufacture of profile coatings in the field of construction components and furniture industry, in a few minutes will be able to reproduce spare parts at their factory, a prototype the size of only two meters.
Uwe Wagner, Managing Director of Düspohl, says:
“Using our machines, our customers in the construction components and furniture industry produce a wide range of different designs and components. At the same time, they must be able to respond quickly and easily to the circumstances of new orders. Therefore, a smart shredding system will be a key component of automation. »
Reduction of grinding processes up to 40%
Fraunhofer IEM researchers began by digitally displaying the grinding process. According to researchers, ” RoboGrinder can eliminate 15% of grinding operations when regrouping from an approximate size to a specified size. The advantage becomes even more obvious when re-sharpening used rollers, where the grinding process takes place directly on the sensitive area, ie re-sharpening should be done with extreme caution so as not to remove too much material. »
Stephen Coppert says:
“With the digital double, we were able to demonstrate that we can eliminate up to 40% of the grinding process. »
Thanks to RoboGrinder: the specified size is observed, the form of a detail is defined by the laser scanner independently and, as additional advantage, very fast automated process of processing
Researchers are currently conducting parametric research and working, among other things, on issues related to construction. The first prototypes of RoboGrinder will be presented at the Hanover Fair in late May.