Context
Kunststof Recycling Van Werven sorts plastic waste. By sorting more and better, both the quantity and quality of the sorted fractions can be increased. Sorting is done manually or automatically: NIR sensors are used to recognise plastic products and separate them by main type: PE, PP or PET. There is a growing demand in the market for higher quality recycled plastics so that high-quality plastic products can also be made from recycled materials. There is also a demand for plastic recyclate with certain guaranteed properties, so that it can be used directly as a substitute for new plastics. This requires increasingly better sorting: plastic subtypes must also be separated from each other, as must products with certain desirable or undesirable properties.
However, the market is increasingly demanding high-quality recycled plastics: materials suitable for high-quality applications and whose properties are reliable and reproducible. This means we need to separate plastic streams even more precisely – down to the subtype level and including specific desired or undesired material properties.
Project Objective
To take this next step, a new generation of recognition applications is needed. Think of: advanced sensors, improved versions of existing detection techniques, combinations of different sensors, and AI-driven recognition that can combine all available signals.
Within this project, sensor manufacturers, research institutions, universities, machine builders, sorting companies, and end users are joining forces. Together, we are working on the accelerated development of plastic sorting 4.0.
Description of the activities
End user DYKA and other Van Werven customers indicate which material specifications are needed to use more recycled plastic in their products. Based on this, research is being conducted into: new sensor technologies, improving existing sensors, combining sensor signals, and developing AI models that recognize complex material properties.
This research is being conducted by Spectral Industries, Bollegraaf, and Wageningen Research, with COAST acting as a sounding board. The new sensors are being built and tested at lab scale. Van Werven will then sort plastics using this technology, after which the recyclate will be used and tested in final products at DYKA.
Van Werven will manage the project; COAST will ensure knowledge sharing and communication. In addition, the project partners will investigate the technical and economic feasibility of sorting systems based on the new sensors.
Results
The project will generate new knowledge about sensor technologies required for plastic sorting 4.0.
The results will be documented in reports and scientific publications. Furthermore, AI models will be developed that operate based on the combined sensor signals.
The technical and economic feasibility of sorting systems using this new technology will also be substantiated and reported.