Is artificial intelligence the solution to all our problems? Or does AI actually cause more problems than it solves? The mood across numerous industrial sectors ranges from gold rush fever to disillusionment. But neither of these reactions does justice to the truth. “Many projects fail because they lack a clear definition of the problem that AI is supposed to solve,” says Paul Lingohr. The researcher at the Chair of Production Engineering of E-Mobility Components (PEM) of RWTH Aachen University leads the “AI-Driven Optimization of Production Processes” consortium study which is scheduled to begin in early 2026 and for which interested companies from all areas of production can still register to participate. The five-to-six-month study promises concrete benefits by discussing real, individual problems in the course of several workshops.
Numerous fields of application
AI technologies are being used more and more frequently in Germanyʼs manufacturing industry – driven in particular by start-up companies and the electrical industry. Having evolved from a niche topic to a strategic priority for competitiveness, AI encompasses much more than the widely used large language models (LLMs) such as ChatGPT or Claude. Artificial intelligence is currently being used profitably in machine monitoring, analytics, and robotics applications. For example, AI-based image evaluation systems are able to automatically detect defects and anomalies in quality control. “However, a lack of expertise in integrating AI into existing processes is still holding back many companies on the path to real efficiency gains,” says digitalization expert Lingohr, who heads PEMʼs “Battery Production Management” research group.
Investments quickly amortized
The organizers of the study emphasize that artificial intelligence reveals significant patterns by recognizing correlations in large amounts of data, which humans are not capable of detecting. The great benefits are particularly evident in production processes with high quality requirements, such as battery production. Even the smallest defects can be identified by AI in real time – far beyond the limits of human perception. This enables a decisive shift from reactive problem solving to proactive error prevention. According to PEM, numerous industry players have already been able to drastically reduce unplanned production downtime and high maintenance costs. Airbus now uses AI-supported data analysis for precise condition predictions for its aircraft, while the energy industry benefits from AI especially in hard-to-access areas such as offshore wind farms. The studyʼs research team claims the return on investment is often achieved after just a few months.
“Data become valuable resources for continuous optimization.”
This then translates into profitability in the form of efficiency gains through flexible production planning and faster changeover times. According to PEM, the greatest multiple benefits lie in longer machine service life, higher quality of intermediate as well as end products, and lower manufacturing costs. “In this way, data become valuable resources for continuous optimization,” Lingohr states.
Goals and content of the study
The aim of the consortium study is to provide participating companies with a comprehensive overview and a sound understanding of the possibilities offered by artificial intelligence. “In many companies, ideas about how AI can be used only scratch the surface of what is feasible,” says Lingohr. In three consecutive workshops, the study will therefore provide helpful recommendations for action based on practical examples and use cases. After gaining a basic understanding of AI and its use in production, participants will work on their own problems in interactive sessions using case studies. To this end, each challenge will first be precisely defined to then demonstrate how it can be solved with the help of specific AI applications. In the third part, a concrete example from research will be used to show how AI solutions can be implemented in industry. As a result, all participants will receive a methodical approach for their individual challenges and AI projects.
Information and registration
Interested companies can apply to participate or clarify any questions by sending an e-mail to p.lingohr@pem.rwth-aachen.de. A maximum of twelve companies can participate in the study.
