Saturn | Showcase end-to-end digitized pultrusion

Pultrusion is a highly efficient production technology for manufacturing fiber-reinforced plastic components. It is particularly suitable for the production of highly stressed components. In reality, however, pultrusion production is currently characterized by a large number of manual work preparations, especially in the setup and management of the process and inconsistent storage of process data. This leads to high costs and time expenditure for manufacturing companies. Also, development work is usually not supported by digital tools. The consequences for manufacturing companies are high and incalculable development costs, making it difficult to enter the market in other sectors.

The Saturn project (Showcase end-to-end digitized pultrusion) is to develop digital solutions for production technology for the manufacture of fiber-reinforced plastic components. In particular, the project focuses on process digitization, production data management, and Intelligent Fiber Guidance using Artificial Intelligence (AI). The solutions offer the potential to reduce setup times by up to 50 percent, improve process control, and accelerate quoting by up to 80 percent. As a result, this manufacturing process's economic use can be realized for a broader range of components.

In the Saturn research project, innovative, digital solutions are being developed in the subject areas listed below:

• Production data management: from data acquisition to data preparation to data analysis (AI).
• Process digitization: local process control at edge level via end-to-end plant control at OPC UA level through smart networking with flexibly interchangeable sub-process components (Industry 4.0).
• Intelligent setup: Intelligent process digitization through knowledge modeling, starting with the automatic design of tools up to optimized and partially automated fiber guidance to reduce the setup effort (Digitization + AI).

The most advanced components are used in all topics, and transferability to related applications and practical benefits are ensured. The solutions incorporate the findings of various research projects, e.g., Semantic Segmentation with CNN from the Air Carbon III project.