Fraunhofer Institute for Systems and Innovation Research ISIBackground
The carbon fiber industry is facing the challenge of establishing a functioning circular economy. One major obstacle is the lack of transparency and networking within the recycling sector. So far, there is no structured digital overview of the relevant stakeholders along the value chain. Fragmented and hard-to-access data sources limit the potential of efficient recycling.
FiberConnect was a subproject of the WIR! Alliance “Wir recyclen Fasern (WIRreFA) (We recycle fibers)”, funded by the German Federal Ministry of Research, Technology and Space (BMFTR) to develop sustainable material recycling for carbon fibers in the Elbtal (Elbe valley) region of Saxony. FiberConnect’s main objective was to develop a digital, AI-supported platform that clearly maps and effectively connects all the relevant stakeholders in the carbon fiber recycling industry in the region, especially new stakeholders or those outside established networks.
Project description
The FiberConnect project provided an overview of regional material recycling and made it possible to close material loops, conserve resources and reduce material waste in the long term.
The project’s goal was to use AI technologies to provide a constantly updated overview of stakeholders along the value chain and to enable better networking within the sector as a result. It employed an AI-supported, largely automated approach: A web crawler continuously searched the internet for relevant companies, and machine learning models used natural language processing to autonomously classify the companies found into predefined categories, such as manufacturers, processors and recyclers. This was used as a basis to generate an always up-to-date overview of the relevant companies in Saxony. These stakeholders were categorized into different phases of the value chain based on their activities, and the results were made publicly available.
