From "Sensing the Sky and Calculating the Earth" to "Calculating the Sky and Calculating the Heavens": Answering the New Test Paper of the 15th Five-Year Plan with "Commercial Aerospace + Artificial Intelligence"

On March 16, the Xiguang-1 06 satellite (Dafu) was successfully launched into orbit and transmitted its first image within 24 hours. As the only commercial aerospace company in China currently possessing full-spectrum hyperspectral remote sensing capabilities ranging from 400 to 2500 nanometers in commercial operational orbit, it is leveraging full-spectrum perception, comprehensive empowerment, and quantitative remote sensing analysis capabilities to support agriculture, forestry, mining, and water conservancy fields. Hyperspectral remote sensing is officially moving from the laboratory to a market worth hundreds of billions of yuan. With "commercial aerospace" being explicitly mentioned again in the draft of the 15th Five-Year Plan and included in the "new industries and new tracks," the entire industry's attention is focused on the same question: What exactly should be done in the next five years?

As a company specializing in hyperspectral remote sensing, Xi'an Institute of Optics and Precision Mechanics CAS Aerospace Science and Technology Group Co.,LTD interprets this national-level "blueprint" not only as a favorable policy environment but also as a profound transformation concerning technological models and industrial logic. If the "14th Five-Year Plan" saw us achieve a technological breakthrough from 0 to 1, then the core theme of the "15th Five-Year Plan" is how to move from "having" to "excellence," building a new ecosystem of "communication, sensing, and computing" that combines "hyperspectral intelligent computing constellation + AI," and achieving a comprehensive leap in large-scale application and intelligent service.

I. The Shift in the "15th Five-Year Plan"From "strategic support" to "industrial pillar"

Comparing the planning recommendations of the 14th Five-Year Plan and the 15th Five-Year Plan, a clear shift in policy focus can be observed. This change signifies that China's aerospace industry is transitioning from a period of technological accumulation to a period of industrial acceleration.

Several key terms in the planning policy serve as beacons for us in the "deep waters":

Scale and Industrialization:The emphasis is no longer solely on "breakthroughs in core technologies," but rather on "enhancing the large-scale production capacity of spacecraft and rocket products" and "developing strategic emerging industry clusters." This indicates that what the nation needs is...Reliable, low-cost, and repeatable industrial supply capacity.

Integration and Servitization:The concept of "integrated space-air-ground communication, navigation, sensing, and computing" heralds the end of the era of single-satellite and single-sensor independent operation. Future competition...It is a competition between systems.,yesComputing power and algorithmsInfiltration in the network.

Application in deep water areas:The plan emphasizes large-scale application in "key industries and mass consumption sectors." This means that remote sensing data must move beyond specialized data centers and become infrastructure like electricity and the internet, empowering various industries.

Under this top-level design, the future of the industry belongs to those who can...Turn data into decision-making, move computing power to space, and lower the barriers to application to the minimum.This is not only an important support for promoting the digital and intelligent transformation of key areas such as agriculture, emergency response, and ecology, but also a core path for China's commercial aerospace to move from "data pipeline" to "value hub" and from "scale expansion" to "intelligent decision-making".

The traditional working mode of remote sensing satellites is "sensing from the sky and calculating from the ground"—the satellite is responsible for taking pictures, and the ground is responsible for interpreting them. This has an inherent weakness in responding to sudden emergencies, as the golden rescue time often slips away during the long data transmission and processing.

The "efficient supply of computing power and algorithms" proposed in the "15th Five-Year Plan" has been given a new interpretation by CAS Xiguang Aerospace: not only should data centers be built on the ground, but computing power should also be deployed in the sky.

This is precisely the core logic behind the "Hyperspectral Intelligent Computing Constellation" being advanced by CAS Xi'an Optics & Aerospace. It pioneers the deployment of "on-board intelligence" driven by hyperspectral applications, using the hyperspectral intelligent computing constellation to deploy distributed computing nodes in space, driving the transformation of remote sensing data processing from "serial processing on-board" to "parallel processing on-board," compressing the traditional "hourly" response time into a "second-level" intelligent closed loop. In September 2025, CAS Xi'an Optics & Aerospace signed a strategic cooperation agreement with the Zhijiang Laboratory, integrating the "Xi'an Optics series" hyperspectral remote sensing constellation into the "Three-Body Computing Constellation" system. The significance of this cooperation is no less than equipping the "eyes" in the sky with a "brain."

The changes brought about by this move arerevolutionaryof.

The revolution of timeliness:Through onboard AI processing, satellites can identify targets and detect changes while still in orbit, directly sending conclusions such as "fire location coordinates" or "contaminated area" to ground users. Emergency response time is reduced from "hours" to "seconds," truly transforming space-based information from "after-the-fact" to "early intervention."

Value chain restructuring:China's hyperspectral remote sensing has officially upgraded from the traditional data return model to...The New Era of "On-Orbit Computing, Real-Time Intelligence" for "Tianshu Tiansuan"We summarize this path as follows:"Heavenly perception, Heavenly calculation, Heavenly application"This is the most direct practice of the "integrated air-space-ground computing network".

If the "Intelligent Computing Constellation" is the hard foundation, solving the problem of timeliness in data acquisition, then the "Hyperspectral Remote Sensing AI Large Model" is the soft engine, solving the problem of depth in data understanding.

The unique value of hyperspectral remote sensing lies in its ability not only to "see" the shape of objects, but also to "see through" their composition by capturing spectral information across hundreds of consecutive bands. Every substance has its unique spectral "fingerprint." However, in the past, identifying these "fingerprints" relied heavily on experience, was inefficient, and difficult to scale.

During the "15th Five-Year Plan" period, it was emphasized that..."Data Value"Against this backdrop, what CAS Xiguang Aerospace is doing is using massive amounts of high-precision spectral data to train large-scale industry models, achieving [the goal of]...From "visible" to "understandable"A leap forward.

In line with policy requirements such as the "Plan for Accelerating the Construction of a Strong Agricultural Nation," the application scenarios of this "soft engine" have great potential.

Precision agriculture: from "looking at the seedlings" to "looking at the ingredients".

Traditional method: Check the crops with a satellite to see if they are green.

Hyperspectral imaging combined with AI can not only identify which plots of land are short of water, but also diagnose crop nutrient deficiencies by retrieving nitrogen, phosphorus, and potassium content; it can even provide early warnings in the early stages of pest outbreaks by detecting subtle variations in leaf spectra, reducing control costs by more than 30%. This is not just a casual "look," but a "quantitative diagnosis" based on material composition.

Ecological and environmental protection and emergency response: From "discovering targets" to "identifying attributes".

Traditional method: An unusual area is found in the river channel, but its nature is unknown.

Hyperspectral imaging combined with AI: In real-time on-orbit identification of whether riverbeds contain ordinary silt or industrial pollution, accurately pinpointing sewage outlets; for forest fires, on-board intelligence can directly penetrate smoke to locate fire points and assess burned area and damage. Related practices by the China Meteorological Administration have also demonstrated that the fusion of hyperspectral and multimodal data enables quantitative monitoring of physiological changes in crops, building a "rapid perception, immediate response" early warning system.

Resource exploration: From "carpet search" to "targeted location".

Traditional method: Geologists trek through mountains and rivers, hammering and drilling.

Hyperspectral + AI: Satellite large-area scanning directly delineates mineralization prospective areas through mineral spectral characteristics, providing an efficient "space-based guide" for energy exploration and mineral prospecting breakthroughs.

The five years of the 15th Five-Year Plan marked a historic leap for China's commercial aerospace industry, transforming it from a "new force" to a "main force." When the nation places aerospace alongside the low-altitude economy as a strategic pillar industry, it not only solemnly affirms the prospects of the aerospace economy but also exerts extreme pressure on the core capabilities of every enterprise. Opportunities always belong to those who master key core technologies, while the challenges are left to those players who have not yet built an industrial moat.

When satellites possess the ability to think, and when the material world is endowed with identifiable digital fingerprints, space information will no longer be a lofty technological concept, but rather an intelligent decision-making process that permeates factory workshops, a precise reach that protects farmland and mines, and an instant response that rushes to emergency sites—this is the ultimate vision of aerospace technology entering ordinary people's homes.

The explorations of CAS Xi'an Optics & Aerospace are a vivid testament to this path. By reconstructing the "brain" of satellites with artificial intelligence and using hyperspectral remote sensing to understand the genes of all things, we are not only developing intelligent computing satellites, but also weaving a space-air information network that perceives the heavens and earth and intelligently computes the future. The onboard computing platform soon to be launched into orbit will be a significant milestone in this vision. There are no shortcuts to becoming a space power; only by taking one step at a time, measuring value with data, and proving strength through applications. "Commercial aerospace + artificial intelligence"—this is not simply a combination of technologies, but a future-oriented industrial transformation, a transformative answer that China's commercial aerospace industry offers to this era.