Fraunhofer IPA and Polysecure build AI robot to sort mixed textile waste
Fraunhofer IPA and Polysecure launched Pick4Crcl on July 1, 2026, an AI robot combining grasping algorithms and multimodal sensors to sort textile waste.
Photo by Freek Wolsink on Pexels
The Fraunhofer Institute for Manufacturing Engineering and Automation (Fraunhofer IPA) in Stuttgart and Karlsruhe-based Polysecure GmbH have started a two-year project to build an AI textile sorting robot capable of separating chaotic mixed piles of clothing and other textile waste by material type, Fraunhofer IPA announced on August 5, 2026. The project, named Pick4Crcl, combines robotic grasping algorithms developed by Fraunhofer IPA with a multimodal material-detection system built by Polysecure, and runs from July 1, 2026, to June 30, 2028, funded through the InvestBW – Innovation IV programme of the German state of Baden-Wurttemberg.
Polysecure is the project’s consortium coordinator; Fraunhofer IPA is the technical partner responsible for the robotic hardware and software. The institute’s press office named Dr. Karin Rohricht as press contact and Florian Jordan as technical contact for Pick4Crcl.
Why an AI textile sorting robot is needed
The world generates more than 92 million tonnes of textile waste every year, according to Fraunhofer IPA, yet less than 1 percent of post-consumer textile sorting in the European Union is currently automated. Most sorting is still done by hand, which is slow and struggles to keep pace with mixed-material garments such as cotton-polyester blends. The pressure to automate has grown since 2025, when EU rules requiring member states to collect textile waste separately from general household waste took effect.
Pick4Crcl addresses two separate technical problems in that chain. Fraunhofer IPA is developing model-free robotic grasping and segmentation algorithms that let a robot arm recognize, pick up, unfold and place individual textile items from a disordered pile onto a conveyor for analysis, without needing to be pre-trained on every possible garment shape. Polysecure is building the material-identification layer: a multimodal classification system based on the company’s patented Sort4Circle technology, which combines several sensor types to distinguish material blends and to detect dark or black-pigmented fabrics that standard optical and near-infrared sensors typically miss. The two systems will be integrated and tested together at a Sort4Circle pilot sorting facility.
The project’s stated performance goals, which are targets for the two-year runtime rather than results already achieved, are a sorting purity above 99 percent for single-material textiles and above 95 percent for mixed-material textiles, with mixed fabrics classified in composition increments of 10 percent. Fraunhofer IPA and Polysecure are aiming for a sorting throughput of up to 60 textile pieces per minute, figures independently confirmed by trade outlet Recycling International, which also reported that Polysecure handles material detection and identification while Fraunhofer IPA handles the robotic separation.
| Pick4Crcl project detail | Figure |
|---|---|
| Project runtime | July 1, 2026 – June 30, 2028 |
| Global textile waste generated per year | Over 92 million tonnes |
| EU post-consumer textile sorting currently automated | Under 1 percent |
| Target purity, single-material textiles | Above 99 percent |
| Target purity, mixed-material textiles | Above 95 percent |
| Target throughput | Up to 60 pieces per minute |
Source: Fraunhofer IPA, “Mehr Textilrecycling dank KI-gestutzter Robotik im Projekt Pick4Crcl,” August 5, 2026; Recycling International, “Can robots unlock better textile recycling?”
As part of the project, the two partners plan to release a public dataset of textile-waste images and sensor recordings, intended to support further AI development for textile recycling beyond Pick4Crcl itself. Fraunhofer IPA has not disclosed a release date or size for the dataset, and no specific euro funding amount for the project has been made public; the institute names InvestBW – Innovation IV, a Baden-Wurttemberg state innovation programme, as the funding source. Winssolutions.org has covered other applications of AI for sustainability across industries, including waste-sorting automation as one of several use cases.
About Fraunhofer IPA
The Fraunhofer Institute for Manufacturing Engineering and Automation (Fraunhofer IPA) is based in Stuttgart, Germany, and forms part of the Fraunhofer-Gesellschaft, the country’s largest organisation for applied research. Fraunhofer IPA specialises in production technology, robotics and automation for manufacturing and recycling applications. The institute’s work on model-free robotic grasping and segmentation algorithms for Pick4Crcl builds on its broader research programme in industrial robotics. With the project running through June 2028, Fraunhofer IPA and Polysecure say the resulting system and its planned open dataset are intended to help push automated sorting beyond the current sub-1-percent level across the EU’s textile waste stream.
Sources: Fraunhofer IPA; Circular Economy News; Recycling International
Featured image: photo by Freek Wolsink on Pexels (free Pexels license).
Become a Sponsor
Our website is the heart of the mission of WINSS – it’s where we share updates, publish research, highlight community impact, and connect with supporters around the world. To keep this essential platform running, updated, and accessible, we rely on the generosity of you, who believe in our work.
We offer the option to sponsor monthly, or just once choosing the amount of your choice. If you run a company, please contact us via info@winssolutions.org.
I specialize in sustainability education, curriculum co-creation, and early-stage project strategy. At WINSS, I craft articles on sustainability, transformative AI, and related topics. When I’m not writing, you’ll find me chasing the perfect sushi roll, exploring cities around the globe, or unwinding with my dog Puffy — the world’s most loyal sidekick.
