November 23, 2022

Trusting artificial intelligence – Metrology and Quality News

Businesses often struggle with artificial intelligence. They are afraid of ending up with a solution that is not always reliable – and therefore often continue to perform tasks, especially complex ones, manually. Fraunhofer IEM researchers have set themselves the goal of countering this skepticism and building trust in artificial intelligence. “We combine the tried and true control technology with the new artificial intelligence,explains Steven Koppert, Head of the Trusted Machine Intelligence Group at Fraunhofer IEM. “While much control technology is 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 exclusively on data and also solves creative tasks. Combine these two approaches and you have a flexible and reliable system, also known as trusted artificial intelligence.

RoboGrinder: an intelligent grinder

Rubber surface laser scanning

What exactly this means can best be explained by the RoboGrinder. At Düspohl, a mechanical engineering company, the process of grinding the rubber rollers that press the films against the component when applying them to window frames or skirting boards is automated for the first time. Until now, these rolls, known as profile wrapping rolls and usually having a complex shape, were ground by hand. Indeed, automation using only control technology is not easily achievable for rubbery material.

Ultimately, it is based on a target/actual comparison and eliminates any excess material. The problem: the roller material is so elastic that it bounces back after grinding – a very difficult behavior to model and integrate into a control technology approach. This means that many grinding processes are required – but the closer the target, the more inefficient the process becomes. In addition, the surface quality may deteriorate with the number of grinding processes.

“We document all grinding errors that occur during each grinding process and use this to train an artificial intelligence model – so that the AI ​​takes over the correction, especially during fine adjustments”, Koppert said. “Comparatively, you could say that the control technology acts as a guardrail, while the artificial intelligence ensures that the curve of the road is negotiated in the best possible way. This way the process can be reliably automated for the first time. For profile coating companies in the component and furniture industry, this means that they can reproduce spare parts in their own factory in minutes. You don’t need a lot of space for this; the prototype measures only one meter by two.

“Using our machines, our customers in the building components and furniture industry produce a wide range of different models and components. At the same time, they must be able to react quickly and easily to the circumstances of new orders. The smart shredding system would therefore be a major part of automation,” says Uwe Wagner, Managing Director of Düspohl.

Reduction of grinding processes by 40%

The grinding process

Fraunhofer IEM researchers first digitally mapped the grinding process. The result: RoboGrinder can eliminate 15% of grinding operations when resharpening from approximate size to specified size. The advantage becomes even more apparent when resharpening used rollers, where the grinding process takes place directly on the sensitive area – meaning that the resharpening must be carried out with extreme caution so as not to remove too much material. “Using the digital twin, we were able to show that we could eliminate up to 40% of the grinding process,” Koppert said. Other advantages of RoboGrinder: the specified size is guaranteed, the shape of the part is detected independently by a laser scanner and the machining process is configured automatically. The setup time is also unique: it only takes a few moments to go from planning to execution.

The researchers are currently conducting parametric studies and are working, among other things, on issues related to construction. The first RoboGrinder prototypes will be at Hannover Messe from May 30 to June 2, 2022 (Hall 5, Stand A06). This should be of interest to all companies that want to automate manual processes reliably, but where this is not possible with standard automation technology. After all, the methodology developed to create trustworthy artificial intelligence is universally applicable.

For more information: