Using AI, scientists have uncovered over 100 new exoplanets and thousands of candidates hidden in NASA data—showing how machine learning is transforming the search for worlds beyond our solar system. (Source: Image by RR)

The Discovery Could Expand NASA’s Exoplanet Catalog Significantly

Scientists have discovered more than 100 previously hidden alien worlds using artificial intelligence, mining years of data from NASA’s TESS (Transiting Exoplanet Survey Satellite). The breakthrough, according to an article in space.com, comes from a new AI system called RAVEN, which validated 118 new exoplanets and identified over 2,000 additional high-quality candidates—nearly half of them entirely new discoveries.

TESS detects planets by measuring tiny dips in starlight when a planet passes in front of its host star. But distinguishing real planets from false signals has long been a challenge. RAVEN streamlines this process by combining detection, analysis, and validation into a single AI pipeline, allowing researchers to process massive datasets—over 2.2 million stars—with greater speed and accuracy than traditional methods.

The findings could significantly expand our understanding of planetary systems. With around 6,000 confirmed exoplanets already cataloged, confirming even a portion of these new candidates would mark a major leap in the search for worlds beyond our solar system. The data also offers new insights into planetary distribution, including the rarity of certain types like Neptune-sized planets in close orbits.

Beyond the discoveries themselves, the research highlights AI’s growing role in astronomy. By identifying subtle patterns in enormous datasets, tools like RAVEN are enabling scientists to uncover phenomena that might otherwise remain hidden—suggesting that the next wave of cosmic discoveries may increasingly come from machines as much as from telescopes.

read more at space.com