A machine-learning–assisted survey of NASA's TESS data analyzed over 83.7 million stars and identified more than 11,000 exoplanet candidates, including 10,052 new candidates that could dramatically expand the known tally.

One confirmation from the Magellan telescope in Chile validated TIC 183374187 b, a hot Jupiter orbiting a star roughly 3,950 light-years away, bolstering the algorithms predictions.

The findings indicate that large-scale, ML-assisted transit searches can significantly expand the exoplanet candidate census, especially around faint stars usually excluded from traditional surveys.

TESS, already yielding 882 confirmed exoplanets, provides the dataset for this work and underscores the potential for rapid inventory growth through advanced data techniques.

While many candidates will need independent verification, the initial confirmation suggests a substantial portion of the new candidates may ultimately be confirmed as real exoplanets.

The study introduced the T16 project, which analyzes transit signals from stars up to 16 magnitudes fainter than typical targets, revealing subtle clues missed in conventional analyses.

Most new candidates have short orbital periods ranging from about half a day to 27 days, indicating many lie too close to their stars to be habitable by Earth-like life.