Machine learning for targeted processing in the recycling of polyesters using reactive extrusion

• Applicant: Professor Dr.-Ing. Holger Ruckdäschel
• Subject Area: Plastics Engineering ; Polymer Materials ; Metal-Cutting and Abrasive Manufacturing Engineering
• Term: since 2023
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 518732456

The proposed project will use digital methods to develop a deeper understanding of the process of (reactive) extrusion of recycled materials. Machine learning (ML) will be enabled based on inline and online recorded process and material data to allow the process to automatically adapt (ideally in real time) to material with input quality. To this end, ML models will first be trained on known materials (PET) and material combinations in increments of increasing complexity and then tested on a post-consumer PET. This should also help make the process more sustainable as material and energy use is quantified and evaluated. Consideration of the remanufacturing process with fresh material and/or chemical modification including digital methods can ultimately be optimized both ecologically and economically.

• DFG Programme: Research Grants