<i>Hubble</i>Asteroid Hunter

Large and publicly available astronomical archives open up new possibilities to search study Solar System objects. However, advanced techniques are required deal with the large amounts of data. These unbiased surveys can be used constrain size distribution minor bodies, which represents a piece puzzle for formation models System. We aim identify asteroids in archival images from ESA Hubble Space Telescope (HST) Science data archive using mining. developed citizen science project on Zooniverse platform, Asteroid Hunter (www.asteroidhunter.org) asking members public asteroid trails HST images. labels provided by volunteers train an automated deep learning model built Google Cloud AutoML Vision explore entire detect crossing field-of-view. report detection 1701 identified via our subsequent machine exploration archive. magnitude 24.5, statistically fainter than populations ground-based surveys. The majority distributed near ecliptic plane, as expected, where we find approximate density 80 per square degree. match 670 (39% found) 454 known objects Minor Planet Center database, however, no matches found 1031 (61%) trails. unidentified faint, being average 1.6 magnitudes succeeded identify. They probably correspond previously unknown This work demonstrates that useful systematic SSOs existing astronomy archives.