Research meets practice: When AI collaborates in production

How can employees be specifically supported in complex production environments? In the DiProLeA research project, SPIE AUTOMATION, together with partners from science and industry, demonstrates how intelligent assistance systems make assembly safer, more efficient, and more capable of learning – an important step towards digitally networked production.

Initial Situation

Production processes are becoming increasingly complex: Products are constantly changing, variants are increasing, and specialists have to adapt to new processes again and again. This is precisely where the DiProLeA research project (Digital Product creation process with learning Assistance system) comes in.

The goal: To digitally connect the entire development and manufacturing process and to specifically support employees through intelligent assistance systems. SPIE AUTOMATION was part of this federally funded project, which was carried out by the Aschaffenburg University of Applied Sciences together with several industrial partners.

Implementation

The project developed a digital assistance system that makes work in assembly more intelligent, safer, and more efficient. Cameras and sensors record the work steps, while an AI in the background recognizes what is happening – and provides assistance if necessary.

For example, the system supports workers with real-time instructions or automatically detects when an error is creeping in. It is constantly learning in order to continuously optimize the processes.

In this development phase, SPIE AUTOMATION contributed not only its research expertise, but also its many years of practical experience in automation, digitization, and software integration. Our colleague Dr. Michael Kröhn, Head of Research & Development, accompanied the project with comprehensive know-how on artificial intelligence, digital processes, and system architectures. Through this combination of innovative research and practice-oriented implementation, we were able to make a decisive contribution to ensuring that the developed technologies are not only convincing in the laboratory, but also function in real production environments and create added value.

Benefits

The practical added value is clear:

Fewer errors because the system automatically recognizes when something is wrong.
Faster training of new employees through understandable, digital support.
Higher quality and more efficient processes because knowledge is stored centrally and can be accessed at any time.

How well theory and practice interact in the DiProLeA project was demonstrated during the practical tests at our project partners from industry: The system was already running successfully in real operation in the assembly of track roller bearings. The XR-based assistance also proved its worth in the manufacture of pipe components in the aviation industry. It not only made the work steps more understandable, but also significantly more precise and efficient.

Our strengths in the project

With our decades of automation experience, we were able to contribute not only technological expertise in the areas of AI, language models, and visual recognition systems, but also our deep practical understanding to the project. In this way, theory became lived practice – an important step towards digitally networked production.