Ideal production thanks to AI
Feintool pioneers the SPAICER research project
Interruptions in production are high risks that may quickly become expensive. In the SPACIER research project, scientists and experts from industry are using artificial intelligence (AI) to develop new solutions that will better protect companies and entire industries from disruptions. Feintool System Parts in Jena is a partner in the SPAICER consortium and is working on smart resilience management – and thus also on business models for the future.
Feintool has been focusing on optimal monitoring of its production processes for years. With SPAICER, the preventive maintenance – so called FEINmonitoring - is now taken to a “completely new level,” explains Jens Gerhard, Head of Technology and Press Development at Feintool System Parts in Jena. The 55-year-old manages Feintool’s stake in SPAICER: pioneering work in close exchange with researchers and experts from other companies. “The potential is huge and the transfer of knowledge exciting every day,” says Gerhard, very pleased. For Feintool customers, the benefits of such an investment in research and development are obvious: process flows and maintenance work will be easier to plan, costs will be reduced, and Feintool’s range of services as a supplier and machine manufacturer will be significantly greater.
The full project name, however, is rather unwieldy: “Scalable adaptive production systems through AI-based resilience optimization.” Put simply, it’s about an experience everyone has had: when incidents arise, it is advantageous to be well prepared and to be able to react appropriately. It is not much different in a globalized production process, where unforeseen disruptions are nothing unusual: they come from outside, from the supply chain, or from a pandemic – and from within, as in tool breakage or machine downtime. In either case, disruptions are a risk to business. After all, an unplanned disruption in the production process can quickly sum up to 500 000 euros or francs an hour.
(Photo by Project SPAICER)
To prevent such a case, not only individual machines or companies but also entire industries must become more resilient. Such an overarching and global approach is new. And that is precisely what makes the research project, which started in April 2020 and is funded by the German federal government with ten million euros over three years, so special. The project is coordinated by the German Research Center for Artificial Intelligence (DFKI) in Saarbrücken. The aim is to use AI and machine learning (ML) to develop digital tools that make it possible to predict events in good time and to adapt quickly and flexibly to changing circumstances. These Smart Resilience Services (SRS) are to be anchored in production networks according to the modular principle. Operations can use SRS via a platform to improve their resilience management.
For the SPAICER concept to work, a lot of data is needed. This is because SPAICER creates digital twins of complete processes. Based on these process simulations, AI methods can be used to develop services that add real value. In total, SPAICER comprises 12 work packages that are closely interlinked. Jens Gerhard’s small team is involved in almost all of them and has designed practical examples (use cases) for various packages.
“With us, it’s tangible just how much AI is taking resilience in production to a whole new level.”
Jens Gerhard, Head of Technology and Press Development at Feintool System Parts, Jena, Germany
“Self-optimization” at the machine level is an example of a use case in which Feintool has the lead. Here, among other subprojects, the focus is on predictions relating to wear. “We want to find the ‘sweet spot’ – that is, the ideal time when a tool is so worn out that maintenance is due, but not a single defective part (N.O.K. part) has been produced yet,” explains Jens Gerhard. To detect wear, data is collected on the tool via structure-borne sound diagnosis. An acoustic procedure – and a whole new field of research. This information can be transformed into concrete recommendations for employees using AI methods. In this way, maintenance arrives with pinpoint accuracy and advance warning, not too early and not too late – without any damage having been done. “The more knowledge we have about the tool’s run time saves time and costs,” summarizes Jens Gerhard.
He also mentions other use cases on a higher or global level that Feintool controls and helps evaluate. It’s always about more and about early knowledge. For example, Feintool wants to build up the steel supplied in coil form as a digital twin and view it in its entirety. Here, too, the technicians receive a lot of new information by means of AI, which, in the end, significantly optimizes the actual fineblanking process and prevents errors.
(Photo by Project SPAICER)
On a global level, the scientists are also looking at climate changes that can have a significant impact on production. Take goods transport, for example: in the summer of 2018, rivers were not navigable at all due to drought. “On this topic, we are looking for new strategies to ensure stress-free procurement, production, and sales by digitizing processes,” explains Gerhard. Exciting Industry 4.0 projects that SPAICER partners from all industries regularly exchange information about in order to learn from each other’s experiences. Knowledge transfer is a big plus for SPAICER. A highlight is scheduled for fall 2021: Feintool will be hosting the 3rd SPAICER consortium meeting in Jena. Feintool will host the third consortium meeting in Jena with research partners and participants from the German Aerospace Center (DLR) and the German Federal Ministry of Economics.
The innovation potential of the research project is enormous. At Feintool, people are thinking far ahead: after all, data is the “gold of tomorrow,” so entirely new business models are emerging in data trading and management. “SPAICER helps us to flexibly design production operations for transformations such as those in the automotive industry. If I can control my processes down to the last detail, I can keep up with changes quickly,” explains Gerhard. This ability to change is the key to sustainable growth. “For this, we need the digitization of the industry. We are working on this future,” he sums up.
Partners Science: DFKI German Research Center for Artificial Intelligence (coordinator), RWTH Aachen University (Laboratory Machine Tools and Production Engineering WZL; Institute for Technology and Innovation Management), University of Freiburg, TU Darmstadt, WHU - Otto Beisheim School of Management
Business partners: deZem, SAP, Schott, Seitec GmbH, Senseering GmbH, more than 40 associated partners