Summary of Key Points
Academician Wei Fengsi from the Chinese Academy of Sciences pointed out in a speech in Shenzhen that the aerospace industry is currently facing challenges such as a surge in the number of satellites and a complex space environment, pushing traditional ground control models to their limits. He proposed a solution called the “Satellite Brain,” which draws on the human brain’s logic of “perception-cognition-action” to endow satellites with autonomous decision-making capabilities. This not only addresses the shortcomings in aerospace intelligence but also has the potential to spawn a new industry worth tens of billions of yuan. If China takes the lead in this development, it could usher in a new era of intelligent aerospace.
The “Six Highs” Challenges of the New Aerospace Era: Too Many Satellites, Too Complex an Environment, and Inadequate Traditional Control
The “six highs” challenges mentioned by Academician Wei Fengsi can be explained as follows:
- High dynamics: Satellite and space environments (such as solar storms and atmospheric density) are constantly changing.
- High timeliness of response: Emergencies (e.g., the risk of satellite collisions) must be addressed immediately.
- High precision control: The adjustment of satellites’ orbits and attitudes requires absolute accuracy.
- High resolution identification: Satellites need to be able to accurately identify other satellites, debris, or unusual phenomena in space.
- Wide coverage: Satellites must cover the Earth and even farther into space.
- Data-driven decision-making: Problems must be solved using data and intelligent technologies.
For example, the 2022 solar storm caused the destruction of SpaceX’s Starlink satellites because they lacked the ability to autonomously perceive and adjust their positions; by the time ground commands were sent, it was already too late. In the future, the number of low-earth orbit satellites is expected to exceed 100,000, and relying solely on ground control would be overwhelming.
Why Give Satellites a “Brain”? Traditional Models Are No Longer Enough
Traditional aerospace control relies on a central ground command center that collects data from satellites, analyzes it, and then issues commands. However, with the increasing number of satellites and the crowded space environment, this approach has significant drawbacks:
- Slow response times: Satellite collisions or other issues may occur before ground instructions can be sent.
- High costs: Ground-based monitoring requires a large amount of personnel and equipment, which becomes more expensive with more satellites.
The “Satellite Brain” is like giving satellites their own “intelligent brains”—they can perceive their environment (e.g., changes in atmospheric density), assess situations (whether they are dangerous), and take action (adjusting orbits or attitudes) without waiting for commands from the ground. This allows for millisecond-level responses and is much more energy-efficient.
The Technical Approach of the Satellite Brain: Learning from the Human Brain, Not Simply Copying Ground AI
The United States has attempted to transfer large-scale models (like ChatGPT) to satellites, but Academician Wei believes this is not feasible due to limited resources in space, such as energy and heat dissipation. Our approach focuses on mimicking the human brain’s causal relationships (e.g., how increased atmospheric density affects satellite behavior). The team has already completed the first step of this three-step process: a ground-based prototype has been successfully verified, capable of processing data and making autonomous decisions. The second step involves conducting experiments on satellites, with plans to test key technologies in low-latitude regions. The third step will involve deploying these systems on space stations or actual satellites to create commercially viable products. The Ministry of Science and Technology is also supporting this initiative, ensuring it is more than just theoretical.
The Potential of the New Billion-Dollar Industry
Academician Wei estimated the potential market size as follows:
- Direct market: If China has 30,000 low-earth orbit satellites in the future and 20% of them are equipped with Satellite Brains, each costing several million yuan, the total value would be in the tens of billions.
- Indirect market: Additional services such as algorithms, data platforms, and maintenance will add significant value.
- Hidden benefits: Reducing satellite failure rates and improving communication/navigation efficiency could lead to even greater savings and value creation.
More importantly, if a global digital aerospace platform can be established to serve all satellites, the industry could grow to tens or even hundreds of billions of yuan in size—similar to how current internet platforms serve users worldwide.
China’s Opportunity: Taking the Lead to Gain Control in Intelligent Aerospace
The digitization and intelligence of the aerospace industry are still in their early stages, and whoever pioneers these technologies will gain a significant advantage. By establishing a Satellite Brain system first, China could independently handle various space-related issues (such as solar storms and collisions) without relying on others’ technologies.
Moreover, “brain-based intelligence” is more suitable for aerospace applications than conventional AI, which requires large amounts of data and high energy consumption to achieve optimal outcomes (e.g., 99% collision avoidance). In contrast, complex systems like satellites require 100% reliability, and brain-based intelligence can achieve this with minimal data and energy usage.
In summary, the Satellite Brain is not only a technical solution but also a new industry driver that could transform the entire aerospace ecosystem. China has the opportunity to lead the world in this field and create substantial economic and technological benefits.