Last week, Gerard Schouten, Leon Schrijvers, Georgiana Manolache and Maurice Hoogsteden-Lamme together with a team of Fontys ICT students traveled to the University of Mons in Belgium, collaborating on an innovative project named Taxobot that aims at assisting zoologists and taxonomists with the precise identification of bee species. The project, grounded in AI technology, seeks to streamline the often challenging task of categorizing European bee species, many of which have subtle physical morphological differences that are difficult to discern without expert knowledge.
During their visit, the researchers and students worked alongside zoologists, closely examining bees under microscopes to understand which physical characteristics are crucial for accurate species identification. By studying the bees’ minute features—such as wing vein patterns, body hair density, and the structure of their antennae the students are now able to use these domain insights and train AI models – in particular the latest LLM and RAG frameworks – to recognize and classify species with precision.
The project’s AI models are being trained on extensive high-quality datasets about bee characteristics, enabling rapid, reliable identification of species. By reducing the time needed for classification, the project aims to provide zoologists and conservationists with the tools they need to monitor bee populations more effectively, ultimately supporting efforts in environmental studies.
The collaboration is expected to advance both environmental science and AI, offering a promising model for future projects dedicated to biodiversity conservation and ecosystem sustainability.