AI system for recording and evaluating waste paper intake and optimizing corrugated base paper production

As part of the digitization of paper production, a project was carried out at Model AG's sites to record and analyze the quality of waste paper using AI.

The quality of the waste paper delivered is recorded at the Weinfelden and Niedergösgen sites by the internal logistics department as the goods are received. In practice, it has so far been difficult to implement a uniform and comprehensive system for recording the quality of the waste paper delivered. Significant fluctuations in the quality of waste paper hamper production, preventing the full potential from being exploited.

UpCircle Technology AG has experience with image capture and AI-supported analysis of waste bales. This technical expertise was applied to the delivery and use of waste paper in a joint project with Model AG.

During the project period from January to July 2025, waste paper deliveries and the loading of pulper belts were almost completely captured and evaluated visually. An AI was trained with the data and is now able to assign batches and individual bales to a DIN 643 grade with high precision. In addition, the proportion of brown paper in the delivery and the individual bale is recorded. In tests on the paper machines, a positive effect in terms of color consumption and strength was achieved with the help of this measured value.

The surface detection of plastic films in waste paper also enables a supplier-related evaluation of disposal costs and can be used as a basis for predictive control of pulper disposal.

In the future, the data obtained will simplify quality checks and can be used for automatic waste paper quality control and complaints management. The waste paper feed to the plants is coordinated with production and should enable improvements in energy, fiber yield, ink, and starch costs.