Portuguese 
English 
Spanish 
3 min of reading 
0 comments 
RAVEN artificial intelligence analyzed data from over 2.2 million stars observed by the TESS mission, confirmed 118 planets, found 31 new worlds, and revealed rare objects, including ultra-fast planets and bodies located in the so-called Neptunian desert.
The artificial intelligence developed by astronomers at the University of Warwick confirmed over 100 exoplanets in data from NASA’s TESS mission, including 31 newly identified worlds. The RAVEN system analyzed signals from stars observed by the space telescope and helped separate possible planets from phenomena capable of mimicking this type of detection.
The research, published in the journal Monthly Notices of the Royal Astronomical Society, used observations of over 2.2 million stars gathered during the first four years of TESS. The work focused on planets very close to their stars, with complete orbits in less than 16 days.
The survey validated 118 new planets and identified over 2,000 high-quality candidates, of which almost a thousand are entirely new. The team considers this set one of the best-characterized samples of planets close to their stars, with the potential to guide future studies.
-
How Chegg lost 14 billion dollars and became a global symbol of the collapse of traditional companies in the face of the relentless rise of generative artificial intelligence
-
Study reveals that being polite to ChatGPT can improve AI responses and avoid surprising failures that affect performance and the quality of interactions.
-
Artificial intelligence will transform 22% of jobs by 2030 and create millions of opportunities while redefining careers, demanding new skills, and revolutionizing the future of global work.
-
An 18-year-old transforms a bedroom-built app into a millionaire machine generating US$1.4 million per month with artificial intelligence and a visionary strategy.
RAVEN expands search for rare and extreme worlds
Among the confirmed planets, there are categories considered especially relevant to astronomy. Some are ultra-short period planets, which complete an orbit around their star in less than 24 hours.
Others are in the so-called Neptunian desert, a region where few planets are expected by current theories. The study also revealed compact systems with multiple planets, including previously unknown pairs orbiting the same star.
The artificial intelligence RAVEN was created to address one of the main challenges in the search for exoplanets: transforming large volumes of data from space telescopes into reliable discoveries. The system looks for small drops in the brightness of stars, caused when a planet passes in front of them.
After this identification, RAVEN uses machine learning models trained with hundreds of thousands of realistic simulations. These simulations include planets and other astrophysical events that can mimic planets, such as eclipsing binary stars.
System evaluates signals and reduces false positives
RAVEN’s strength lies in its complete processing of the analysis workflow. The tool detects the signal, evaluates its origin with machine learning, and performs statistical validation of the strongest candidates.
This operation differentiates the system from tools that only act on specific parts of the process. The proposal is to analyze enormous datasets consistently, objectively, and sufficiently validated for use in population studies.
In addition to accelerating the discovery of new worlds, artificial intelligence also measures which types of planets are easier or harder to find. This step helps correct hidden biases in the data and allows for cleaner samples to answer broader questions about the frequency of different planets in the galaxy.
Study measures frequency of nearby planets
With the validated baseline, researchers moved beyond individual discoveries. In a complementary study, also published in MNRAS, the team measured the occurrence of nearby planets around Sun-like stars.
The results indicate that about 9% to 10% of Sun-like stars host a planet in a close orbit. This result aligns with previous measurements from the Kepler mission, but the new analysis reduced uncertainties by up to ten times.
The work also presented the first direct measurement of the rarity of planets in the Neptunian desert. These objects appear around only 0.08% of Sun-like stars, reinforcing the unusual nature of this region.
Catalogs become available for new observations
The team released interactive catalogs and tools for other scientists to explore the results. These materials can help select promising targets for follow-up observations with ground-based telescopes and future missions, such as the European Space Agency’s PLATO.
Studies show how artificial intelligence is expanding the analytical capacity in astronomy. By combining large databases with machine learning, RAVEN found hidden planets, validated candidates, and produced a sample capable of more precisely measuring the presence of worlds near Sun-like stars.
With information from ScienceDaily
Be the first to react!