Artificial intelligence is already providing solid support in astronomical data collection, analysis and processing, as well as in satellite engineering and system management, and is expected to further assist astronomers worldwide in their search for exoplanets and space exploration, experts said in Shanghai on Thursday.

Zhang Yonghe, deputy head of the Chinese Academy of Sciences' Innovation Academy for Microsatellites in Shanghai, was among those who spoke during the 2024 World Artificial Intelligence Conference, which ran from Thursday to Saturday.

"Astronomy is about discovering the unknown, and AI can help in various aspects when it comes to astronomy," said Zhang, who is also director of the Key Laboratory of Spacecraft Digitalization Technologies, and lead designer of the Einstein Probe and chief commander of the Space Variable Objects Monitor, both of which are telescope satellites.

He said AI can help make a big difference in data analysis and processing.

Though many people may be familiar with megabytes, gigabytes and terabytes, astronomy generally measures data in petabytes. One petabyte is equal to about 1,000 TB.

"Each astronomical satellite can generate massive amounts of data, and the quantity can be too much for the human brain to process," Zhang said.

"Also, astronomical data is complicated. Some data can be extremely subtle, but it often indicates unknown phenomena and new discoveries, especially in the search for exoplanets," he said, adding that AI can help astronomers identify accidental discoveries.

In addition to data processing, Zhang said AI could also play a big role in spacecraft development and the management of observational satellites.

For example, AI can accelerate the satellite design process by swiftly creating plans for an optimal structure.

"An astronomical spacecraft usually requires much more advanced technologies than existing ones when it is initially proposed," Zhang said. "That is to say, we have to think about the future and carry out error analysis to find out the right way of doing things. And this is where AI is key — it can help us face technical challenges at much lower costs."

In recent years, China has seen rapid development in space science. The nation's first astronomical satellite, China's Dark Matter Particle Explorer — nicknamed Wukong, or "Monkey King" — was launched in December 2015 and is still functioning well.

In August 2016, China launched the world's first satellite dedicated to testing the fundamentals of quantum communication in space. Six years later, the Advanced Space-based Solar Observatory, a comprehensive solar observation satellite dubbed Kuafu 1 in Chinese, was sent into space and has detected more than 100 solar white-light flares, according to CAS' Purple Mountain Observatory.

On Jan 9, the Einstein Probe was sent into orbit from the Xichang Satellite Launch Center in Sichuan province to observe astronomical phenomena. Last month, the Space Variable Objects Monitor, jointly developed by China and France for two decades, was launched to capture and observe gamma-ray bursts.

"With continuously increasing investment in the astronomy and space sector, China has been making efforts to increase the number of astronomical satellites in recent years," Zhang said. "As more satellites are produced, the human brain is less capable of efficiently managing aspects such as their operations and data transmissions."

With the rapid development of AI systems, astronomy is advancing faster than ever before.

"Astronomy involves global collaboration and coordination," Zhang said. "When one spacecraft observes a phenomenon with high value, it will promptly transmit the information to other observation devices across the world, including other spacecraft in space and telescopes on Earth. Given that most of the devices are unattended in space, artificial intelligence can play a major part in the data-sharing process.

"Any spacecraft or device has a limited view on its own, but others located in different places can be complementary. Hence, the more astronomical spacecraft there are, and the more intelligent they are, the more opportunities for new discoveries will arise."

[email protected]