Airborne bioparticles, notably fungal spores, pose health risks, necessitating precise monitoring. While manual methods exist, interest is shifting towards automated systems employing machine learning. However, challenges persist due to diverse particle properties and limited training data. This study, part of SYLVA and COST Action ADOPT, addresses these gaps by establishing best practices for cultivating reference material and creating tailored datasets. Seventeen fungal species were tested on Plair RapidE+ and SwisensPoleno Jupiter. Proof-of-principle models using holography and fluorescence data were developed, achieving variable genus classification accuracy (0.43-0.95). The protocol enhances automatic identification of airborne fungal spores, promising more efficient monitoring systems.