Summary of Key Points
This article highlights the disruptive impact of AI on the aerospace industry: leading international companies such as SpaceX and Palantir are using AI to transform how aerospace systems are operated, data is utilized, and business models are designed. They are shifting from focusing on hardware production to defining intelligent solutions. In contrast, domestic commercial aerospace firms remain stuck in traditional hardware-centric approaches, facing critical challenges such as manually maintaining large constellations of satellites. AI is not merely a supplementary element; it represents a game-changer that will redefine the value of the aerospace sector, with a focus on software and AI-driven technologies.
1. Satellite Operations: A Paradigm Shift from “Nokia” to “iOS”
Traditional aerospace operations are akin to the “feature phone era”: each satellite requires its own dedicated team for management, with calculations taking three days, verification another two days, and execution one day. This approach is feasible for a small number of satellites, but becomes impractical for tens of thousands—consider China’s ambitious “Ten Thousand Stars” constellation, which may require tens of thousands of people to maintain manually.
SpaceX’s Starlink system, on the other hand, operates like a smartphone: each satellite is equipped with an AI chip that enables autonomous collision avoidance. These satellites receive real-time orbital data and make their own decisions about potential collisions (adjusting their orbit automatically if the risk exceeds one in a hundred thousand). This not only represents a technological advancement but also transforms satellites from cold industrial products into self-capable intelligent entities.
2. Space Data: A Business Model Revolution from “selling Images” to “Providing Insights”
Traditional remote sensing satellite companies earn money by selling raw images, while Palantir focuses on providing actionable insights. For example, upon detecting an anomaly at a military base, AI automatically schedules the nearest satellite to take a photo and identifies the location in real time (e.g., “An launch vehicle is located at coordinates XXX”), directly delivering the information to commanders. This significantly reduces the time required from anomaly detection to intelligence generation—from days to minutes.
This shift represents a higher value addition in the supply chain, as customers are no longer interested in just the data but in actionable recommendations on how to proceed.
3. New Paths for Small Companies in AI-Aerospace
Even beyond giants, small companies are exploring innovative approaches:
- Loft Orbital: A “Android” for the satellite industry, providing a standardized platform that allows customers to use space computing power without having to build their own satellites. They simply upload AI algorithms to the satellites, reducing the development time from years to days.
- Slingshot: Similar to Google Maps in space, this company uses AI to manage satellite traffic, identifying whether satellite movements are normal or suspicious (e.g., whether they are being deliberately approached by foreign satellites). Its technology was adopted by the U.S. Space Force in 2023.
- Capella&ICEYE: Their AI technology enhances the effectiveness of SAR satellites by removing noise from cloud-covered images and accurately identifying targets (e.g., the type of ships at a port), while only transmitting data from abnormal areas, thus reducing costs.
4. Challenges Faced by Domestic Aerospace
The article highlights several critical issues facing domestic commercial aerospace firms:
1. The Need for Collaboration: Without forming alliances, companies risk being left behind. Palantir does not manufacture satellites but uses AI to manage entire constellations. There is a lack of cooperation between domestic AI and aerospace companies; large AI models often do not understand the constraints of the aerospace industry, and aerospace firms are unsure of what AI can achieve.
2. The Importance of Onboard Computing Power: In the future, satellites will be evaluated based on their computing speed rather than image resolution. Companies like Loft Orbital are incorporating Google’s TPU chips into their satellites. Radiation-resistant AI chips and onboard operating systems will become critical factors.
3. **The New Meaning of “Autonomy and Control”: As foreign companies set industry standards using software, domestic firms are still debating the extent of localization. The owner of the interfaces determines who controls the ecosystem—this is not an exaggeration.
Conclusion: It’s a Matter of Survival, Not Technological Choice
The article reiterates the main point: when foreign companies use AI to transform aerospace into a software-driven field, domestic firms that continue to focus on hardware production are on the wrong path. Time waits for no one in space—debris orbits the Earth at 7 kilometers per second. By the time domestic companies decide whether to adopt AI, the technology will have already made significant progress. AI is not an optional feature; it is a necessity for survival in the aerospace industry.