This research was conducted by the University of Warwick in the United Kingdom, and its results were published in three separate articles in the journal Monthly Notices of the Royal Astronomical Society. Known as the “hunter for alien worlds,” the TESS telescope measures tiny changes in the star’s brightness as a planet passes in front of its star (transit). The researchers had the data obtained from these observations analyzed by the artificial intelligence system called Raven. The system examined more than 2.2 million observations from TESS’s first four years of data, focusing particularly on planets that orbit very close to their stars, meaning they orbit in less than 16 days.
Marina Lafarga Magro, one of the authors of the research, stated that thanks to Raven, 118 exoplanets were confirmed and more than 2 thousand candidate planets were detected. Approximately 1,000 of these candidates are completely new discoveries. Magro stated that these data are one of the best examples yet of short-orbit planets and will help identify the most promising planetary systems in the future.
The remarkable results of the study include planets that orbit their stars in less than 24 hours, planets located in regions called “Neptune deserts” where Neptune-like planets are expected to be rare, and systems containing more than one planet in orbits close to their stars. Additionally, analysis revealed that approximately 10% of Sun-like stars have at least one planet in close orbit, while Neptune-like planets are found in only 0.08% of these stars.
The Raven system helps distinguish whether a decrease in a star’s brightness is actually caused by a planet or another factor. Andreas Hadjigeorghiou, one of the developers of the project, stated that machine learning models are trained to recognize patterns in data. According to researchers, artificial intelligence has not only discovered new planets; It also revealed a reliable tool for mapping the distribution of different types of planets around Sun-like stars.
