Chinese scientists and engineers are developing a space-based computing network that will enable satellites to process data in orbit in real time, paving the way for supercomputing capabilities in space.
The initiative, led by the Chinese Academy of Sciences' Aerospace Information Research Institute, took a concrete step earlier this week with the establishment of a joint venture, Xingyun Space AI Technologies, in Hefei, Anhui province. The joint venture, backed by companies invested in by the academy and other State-owned investors, aims to integrate orbital computing resources with ground-based infrastructure.
Wu Yirong, academic dean of the CAS institute, said at an industry forum in Hefei on Monday that the traditional satellite operating model — in which satellites collect imagery and other data in space, transmit raw files back to Earth and rely on ground-based computers for processing — suffers from significant delays and poor timeliness.
For example, data collected over oceans or related to general aviation and commercial drone operations often cannot be transmitted to ground stations promptly, delaying analysis and decision-making, he said.
"Space-based computing addresses this problem by equipping satellites with integrated sensing, computing and communication capabilities, enabling data to be processed as it is collected. Connected through high-speed inter-satellite links, these orbital computing resources can form a complementary network to ground-based computing systems," Wu said.
Shao Zongyou, chairman of Xingyun Space AI Technologies, said at the forum that space-based computing involves far more than simply transferring ground-based computing hardware to satellites.
"Instead, it is a complex undertaking spanning remote sensing, semiconductors, space networks, spacecraft thermal management, energy supply and constellation control. Whoever builds the first shareable, maneuverable and operable computing network in space will gain the lead in the space-based information industry," Shao said.
He said the company's planned "space cloud" model will integrate remote-sensing satellites, communication satellites and dedicated computing satellites into an intelligent space-based information service system, covering the entire industrial chain, from domestically developed computing chips to end-user data applications.
Xingyun Space AI Technologies plans to gradually deploy three categories of satellites: Remote-sensing satellites equipped with high-performance edge computing capabilities; general-purpose computing satellites with substantial processing capacity; and supercomputing satellites designed to deliver ultrahigh computing performance.
Once established, the network will enable the seamless integration and unified scheduling of computing resources in space and on the ground, allowing users to access orbital computing services on demand, Shao said.
When fully operational, the infrastructure will form a space-based information system integrating real-time sensing, high-throughput communications and high-performance computing.
It will serve public-sector needs, including natural resource surveys, emergency response and disaster relief, as well as low-altitude aviation operations. It will also be connected to the national computing internet system to provide integrated space-ground computing services to users worldwide.
Huang Cong, a senior researcher at the National Satellite Meteorological Center in Beijing, said that equipping satellites with onboard computing and data-processing capabilities would be highly significant.
China will deploy ultralarge networks comprising thousands of spacecraft, Huang said, adding that satellites in such constellations will operate in relatively close proximity and need to maintain their orbital positions to keep formation.
"When space weather conditions change, transmitting adjustment commands from ground stations via uplinks is too slow to ensure a timely response. If the network has built-in onboard computing capabilities, allowing satellites to process data in orbit and make autonomous adjustments, they can automatically adjust their attitude and position to mitigate potential risks," he said.
Another potential application is disaster response. Satellites equipped with computing capabilities can process data in orbit in real time and generate images and analytical results of affected areas, providing time for search and rescue operations, Huang said.
