Astronomers at the University of Warwick have reached a major milestone in space exploration by using a custom artificial intelligence tool to identify over 100 previously "hidden" exoplanets.

Axar.az reports that using a newly developed AI pipeline called RAVEN, the team analyzed data from NASA’s Transiting Exoplanet Survey Satellite (TESS), scanning over 2.2 million stars to detect the subtle dimming of starlight caused by passing planets.

This breakthrough has already led to the validation of 118 new planets and the identification of over 2,000 high-quality candidates.

The RAVEN system is designed to solve a major bottleneck in modern astronomy: distinguishing real planets from "false positives" like eclipsing binary stars or instrumental noise.

The study particularly focused on "close-in" planets that orbit their stars in less than 16 days. Among the discoveries are rare "ultra-short-period" planets that complete an orbit in under 24 hours, as well as elusive worlds in the "Neptunian desert," a region where planets were theoretically thought to be scarce.

Dr. Marina Lafarga Magro, the study's first author, noted the significance of the find: “Using our newly developed RAVEN pipeline, we were able to validate 118 new planets... This represents one of the best characterised samples of close in planets and will help us identify the most promising systems for future study.”

Beyond individual discoveries, the AI has provided unprecedented precision in mapping planetary populations.

The team found that approximately 9–10% of Sun-like stars host a close-in planet, a figure that matches previous NASA Kepler mission data but with ten times less uncertainty.