According to new study from the University of Georgia, artificial intelligence may be used to locate planets beyond our solar system. The latest research proved that machine learning can be used to locate exoplanets, which might change the way astronomers detect and identify new planets distant from Earth.
“One of the interesting aspects of this is evaluating situations where planets are still developing,” said Jason Terry, primary author of the research and PhD student in the UGA Franklin College of Arts and Sciences department of physics and astronomy. “Previously, machine learning was seldom used to the sort of data we’re utilising, particularly when looking at systems that are still actively creating planets.”
The first exoplanet was discovered in 1992, and although over 5,000 are known to exist, they were among the simplest for scientists to discover. Exoplanets in their early stages are difficult to detect for two reasons. They are too far away, frequently hundreds of light years from Earth, and the discs that develop around them are extraordinarily thick, thicker than the distance between the Earth and the sun. According to the data, planets tend to reside at the centre of these discs, leaving a signal of dust and gases thrown up by the planet.
The study demonstrated that artificial intelligence may assist scientists in overcoming these challenges.
“This is a really exciting proof of concept,” said Cassandra Hall, associate professor of astrophysics and chief investigator of the Exoplanet and Planet Formation Research Group. “The power here is that we trained our AI entirely on synthetic telescope data created by computer simulations before applying it to actual telescope data. This has never been done before in our discipline, and it lays the stage for a flood of discoveries once data from the James Webb Telescope becomes available.”
NASA’s James Webb Space Telescope, which was launched in 2021, has started a new level of infrared astronomy, offering breathtaking new photos and reams of data for scientists to evaluate. It’s simply the latest iteration of the agency’s search for exoplanets, which are spread around the cosmos.
The Nancy Grace Roman Observatory, a 2.4-meter survey telescope that will seek for dark energy and exoplanets and is expected to debut in 2027, will be the next big increase in capability—and transmission of information and data—to comb the cosmos for life.
The Webb telescope allows scientists to see exoplanetary systems in exceptionally bright, high-resolution images, with the developing environments themselves of major importance as they shape the resulting solar system.
“The potential for useful data is growing,” Terry remarked.
New analytical tools are required.
Next-generation analytical tools are desperately required to deal with this high-quality data so that scientists may spend more time on theoretical interpretations rather than manually going through the data looking for tiny small signals.
“In a way, we’ve simply produced a better person,” Terry said. “To a large extent, we analyse this data by looking through dozens, hundreds of images for a specific disc and asking ‘is that a wiggle?’ then running a dozen simulations to see if that’s a wiggle and… it’s easy to overlook them—they’re really tiny, and it depends on the cleaning, and so this method is one, really fast, and two, its accuracy gets planets that humans would miss.”
According to Terry, machine learning may already boost human ability to save time and money while also effectively guiding research time, investments, and new suggestions.
“There is still scepticism about machine learning and AI in science, especially in astronomy, a reasonable critique of it being this black box—where you have hundreds of millions of factors and somehow you get an answer. But we believe we’ve shown in this study that machine learning is capable of the job. You may debate interpretation. In this situation, however, we have very tangible findings that illustrate the effectiveness of this strategy.”
The effort of the research team is intended to provide a solid basis for future applications on observational data by showing the method’s usefulness using simulational observations.
You might also be interested in reading, A model that might help robots grip items better.
