_edited.png){kind=small}
INOV researchers tell us about the process of developing a state-of-the-art, portable spectrometer to analyse cultural heritage objects' spectral fluorescence signatures.
The preservation of cultural heritage is not merely a historical concern—it is an urgent, contemporary challenge that demands innovative solutions. At INOV, researchers are deeply committed to bridging the gap between technology and heritage conservation, a mission exemplified by our contribution to the Horizon Europe ANCHISE project.
As part of ANCHISE, INOV has developed a state-of-the-art, portable spectrometer to analyse cultural heritage objects' spectral fluorescence signatures. This means we illuminate the samples with weak light and then measure the light coming back from the object. This technology enables dating ancient books by examining their unique fluorescence emissions. Crucially, this device incorporates machine learning algorithms to enhance its accuracy and efficiency, providing a reliable tool for combating illicit trafficking. We are starting with books from which we know the date to be able to train our machine learning algorithms. In the future, we hope to be able to apply our method to measuring paper samples from unknown dates.
Our work in the ANCHISE project aims to equip cultural heritage professionals—archaeologists, museum curators, and law enforcement agencies—with the technological tools to protect Europe's cultural heritage. By leveraging advanced spectrometry and machine learning, we contribute to a larger effort to ensure these priceless artefacts remain safeguarded for future generations.
The integration of machine learning enhances the spectrometer data analysis by allowing learning from each sample. In fact, our human perception does not allow us to find the features relevant for dating the paper from books. No easily understandable correlation exists between the peaks and valleys in fluorescence spectra and book date. In addition, the capability of analysing the input data with machine learning, without human intervention, is vital when creating a simple-to-use tool.
The use of machine learning would not have been possible without the data gathered in the libraries of Instituto Superior Técnico from Universidade de Lisboa and in the Academia das Ciências de Lisboa. The generous contribution of these two institutions was much appreciated.
We have already measured more than 300 books; the more, the better, without harming the books. The sampling must be done carefully because the paper composition has evolved over the centuries, from cotton of rags to wood pulp, and the pulping methods have also varied.
In September 2024, we presented our innovative measurement method during the first ANCHISE project demonstration in Thessaloniki, Greece (NDR: find here the recap of the first ANCHISE demo). This event brought together experts from across Europe to witness the potential of the tools developed within the ANCHISE consortium. The feedback we received from this demonstration is invaluable, as it will help us further refine the spectrometer to meet the specific needs of professionals in the cultural heritage protection field.
As we continue to develop and improve this technology, we remain confident that the spectrometer, alongside the other tools in the ANCHISE project, will play a pivotal role in the global fight against the illegal trafficking of cultural goods.