🇩🇪 STUTTGART, Germany — Researchers in Germany are developing an AI-powered robotic system designed to automate one of the most challenging steps in textile recycling: separating and identifying mixed post-consumer clothing before it can be recycled.

♻️ The Pick4Crcl project, led by the Fraunhofer IPA together with Polysecure GmbH, combines AI-based robotic grasping with multimodal sensor technology to recognize, pick, unfold and classify textiles for material analysis. The system is designed to process chaotic textile streams, including dark fabrics, blended materials and tangled garments that remain difficult to sort using conventional automation.

📊 The technology addresses a growing challenge for the recycling industry. More than 92 million tonnes of textile waste are generated worldwide each year, while less than 1% of textile sorting in the EU is currently automated. Demand for advanced sorting solutions has also increased since the EU introduced mandatory separate collection of textile waste in 2025.

Beyond the robotic system, the project will create a publicly available dataset of textile waste images and sensor data to support future AI development and accelerate the deployment of automated textile recycling technologies.

Image: Fraunhofer IPA