The real impact of FasTEX lies in creating open research infrastructure. By placing both the spectral data and the analytical code in the public domain, we are providing a common foundation that researchers can benchmark against, reuse and expand. We hope this approach inspires wider data sharing in the field, helping to reduce duplication of effort and speed up innovation across environmental and forensic fibre research.”
Dr Matteo Gallidabino, Lecturer in Forensic Chemistry, and senior author of the data article
04 March 2026
Researchers create open 'fingerprint library' to identify textile fibres
Researchers at King’s College London have developed an openly accessible reference library of infrared “fingerprints” that can support rapid identification of textile fibres using machine learning.
As part of the project, the team has published a data article in Data in Brief describing the dataset structure, accompanying benchmarking analyses, and the open research resources that support reuse of the data.
The FasTEX project addresses a growing challenge across environmental science, forensic investigation and textile sustainability: reliable identification of textile fibre composition. Textile-derived fibres, including microfibres, are now widely detected in water, soil and air, yet identification can be slow, labour-intensive and heavily dependent on specialist expertise.
To address this, the researchers curated a high-quality dataset of infrared spectral data designed to modernise fibre identification, making it faster, more reproducible and more scalable.
Building a shared reference for fibre identification
Using infrared spectroscopy, the researchers examined 137 textile samples of known composition under standardised laboratory conditions, creating a library of chemical “fingerprints” for each fibre. The dataset includes 26 fibre types, spanning both natural and man-made materials.
The data have been organised for reuse by other researchers, alongside open tools that allow the analyses to be tested and reproduced. The team also explored how machine learning can be used to automatically identify fibres from their infrared signatures.
Benchmarking results indicate that combining infrared analysis with machine learning can support accurate and consistent textile fibre classification within the scope of the dataset.
Environmental, forensic and open-science impact
Accurate fibre identification is essential for understanding microfibre pollution, tracing pollution sources and improving recycling and circular fashion systems. Existing reference libraries are often incomplete, inconsistent or inaccessible.
The FasTEX dataset helps address an analytical bottleneck by enabling faster and more objective fibre identification. It also supports ongoing efforts to improve objectivity and reproducibility in forensic fibre analysis, where fibres have long played a role in criminal investigations.
The project also establishes open research infrastructure for fibre analysis, rather than individual research groups building isolated reference collections.
To the team’s knowledge, this is among the first openly available textile fibre datasets of this kind to combine raw measurements, processed reference data and analysis code in one place.
The project team is also developing a dedicated website to make the spectral data easier to browse and use for non-specialists.
The FasTEX project was supported by a Research Development Grant from the IMPACT+ Network, part of the UKRI Circular Fashion and Textiles Programme funded by NERC.
LINKS
- Data repository (Mendeley Data): https://data.mendeley.com/datasets/rx3fjgz96x/3
- Machine-learning code (GitHub): https://github.com/reeha-parkar/FasTEX-KCL
In this story
Lecturer in Forensic Chemistry