This project will provide the basis for a commercial, high quality and expansive Ground Truth Dataset for textile fibres to enable AI-powered fibre analysis to deliver step change in the UK and Global markets for textile recycling. We anticipate the resulting dataset to serve as a cornerstone for advancing the UK and Global market for textile recycling by improving sorting accuracy, enhancing recycling efficiency, promoting circular economy principles, and fostering innovation. By harnessing the power of data-driven solutions, the UK can build a more sustainable and resilient textile recycling ecosystem, contributing to both environmental conservation and economic prosperity.

Using advanced deep learning, computer vision models, multispectral sensors & automated robotics to develop an end-to-end UK recycling framework for waste garments.

**PORTABLE AiFibres - Using advanced deep learning, computer vision models, multispectral sensors & automated robotics to develop portable, easy to use scanning and sorting system for waste garments and fabrics.**

AI-powered cloth recycling reduces the environmental impact of the fashion industry by reusing and recycling textiles, minimizing resource depletion, water pollution, and greenhouse gas emissions. Moreover, it helps reduce textile waste, which is a significant global concern. AI technologies can effectively sort and process textile waste, increasing recycling rates and reduce landfills. Also, it conserves valuable resources like water, energy, and raw materials by reusing fibers instead of producing new textiles from scratch minimizing reliance on virgin resources. Additionally, it creates economic opportunities by establishing new industries and job roles in textile recycling. It also drives innovation and technological advancement by combining AI algorithms, machine learning, and robotics. It has been claimed that one of the most difficult issues in the automated textile recycling process is the removal of zips and buttons. As a result, textile waste must be scanned and sorted in accordance with the applicable user requirements. Our main areas of focus is to design, develop and build an automated AI-powered textile sorting machine that makes it possible to handle large textile flows and detect textile composition (cotton, poly, etc.) percentages that are better adapted to different recycling processes than today's manual sorting. The proposed project is to develop and create an AI-enabled semi-automated textiles recycling framework, as well as to provide lab demonstrations accompanied by a small-scale machine prototype that can detect and classify clothing based on their compositions and remove undesired features such as zippers. As part of the AI4FIBRES project, many use cases and pilots will be planned and conducted to demonstrate the proposed framework and machine's potential to handle up to 10 tonnes of used clothing each week. The machine learning and computer vision system for textile analysis, classification, and segmentation using a multispectral camera-based architecture, will be integrated into our proposed platform. The overall purpose is to further automate the process of textile analysis reducing the expected processing time per unit, the overall cost and increase the volume of processing items. As a result the impact to many families, local communities or boroughs and the general environmental impact is high and significant. AI4FIBRES will offer support on designing and evaluating advanced novel AI and computer vision solution for textile and fabric classification reducing the errors. Additionally data will be collected using hyperspectral sensors aiming to improve the AI solutions and offer an automated textile processing platform.

Globally around 14.3 million tons of textile waste is sent to landfill each year. This is nearly 5% of all landfill space. In the 2019 WRAP report, it estimates 350,000 tons of clothing goes into landfill in the UK which is the same as each person throwing away more than 5 kg of their clothing per year. Our project proposes to use waste textile garments to create fibrous insulation for applications like high performance outerwear by utilising a fully sustainable natural binder.