At 10:54 on July 22, 2026, the first domestically developed quantitative hyperspectral intelligent computing satellite, "Xiguang-2 01 Satellite (Caiyun Hyperspectral 01 Satellite)", independently developed by Xi'an Zhongkesi Aerospace Technology Group Co., Ltd. (hereinafter referred to as "Zhongkesi Aerospace"), was successfully launched into orbit in the East China Sea using the Gravity 1 Yao-4 carrier rocket. All telemetry parameters were normal, and the in orbit state was stable and good. The launch mission was a complete success. This satellite is the first satellite customized and developed by China Aerospace Science and Technology Corporation for Yunnan Geological and Mineral Exploration and Development Bureau (hereinafter referred to as "Yunnan Geology and Mineral Resources"), and it is also the first hyperspectral satellite in the Yunnan Geology and Mineral Resources "Color Cloud Constellation" series. At the same time, it is also the first AI computing satellite that integrates breakthroughs in hyperspectral quantitative remote sensing technology and real-time intelligent processing in orbit after the strategic cooperation between China Aerospace Science and Technology Corporation and Zhijiang Laboratory.
Relying on the advantages of the new interference system, the star has achieved the highest commercial spatial resolution and maximum bandwidth in the visible near infrared spectral range, as well as the highest commercial hyperspectral indicators of 10m spatial resolution and 7nm spectral resolution in China, greatly improving the country's ability to obtain spaceborne hyperspectral data. Under this indicator, the highest width in China is 14km, which also has the ability to continuously scan and image long strips. The data acquisition efficiency is greatly improved, and up to 1 million km ² of 10m hyperspectral remote sensing images can be obtained per day. Adopting a new system of interferometric spectroscopic imaging technology, the center wavelength can be freely defined, providing a solid data foundation for quantitative and refined remote sensing applications.
As the first hyperspectral satellite in the domestic commercial aerospace field to adopt interferometric spectral imaging system, this satellite breaks through the traditional data processing mode of "sky sensing and earth computing", directly completes the acquisition, processing and decision-making of remote sensing data in orbit, and takes the lead in achieving a historic leap in the field of hyperspectral remote sensing with "day sky computing", possessing the dual breakthrough capability of "quantitative remote sensing+real-time response" for the first time in China.
In the future, the satellite will focus on serving mineral resource exploration, ecological monitoring of mining areas, and full lifecycle management of minerals in Yunnan and globally, injecting space-based energy into the green transformation and digital governance of the mining industry. The successful orbit insertion of Caiyun Hyperspectral 01 satellite marks a crucial step forward in China's fully autonomous and controllable quantitative hyperspectral intelligent computing system, and also sets a new benchmark for higher technical indicators, more representative applications, and stronger intelligent processing in China's commercial aerospace remote sensing field.
Figure 1: Xiguang-2 01 satellite (Caiyun Hyperspectral 01 satellite) launched into space aboard the Gravity 1 Yao-4 carrier rocket
With the successful launch of the Caiyun Hyperspectral 01 satellite into orbit, the total number of satellites in orbit for China Aerospace Science and Technology Corporation has exceeded 12.
Building a solid foundation: Triple technological breakthroughs reshape space-based perception capabilities
According to Hu Xin, the technical director of Caiyun Hyperspectral 01 satellite, the satellite is equipped with hyperspectral optical payload, panchromatic optical payload, and in orbit intelligent processing payload, which solves the pain points of traditional remote sensing "sky sensing and earth computing" such as time lag and insufficient recognition accuracy, and achieves three hardware level technological breakthroughs, laying the foundation for the universal technical standard of "Xiguang series" quantitative hyperspectral intelligent computing constellation single machine.
Figure 2: R&D personnel debug Xiguang-2 01 star (Caiyun Hyperspectral 01 star)
Firstly, interferometric spectroscopic imaging: from "qualitative remote sensing" to "quantitative analysis".
The star adopts an interferometric spectral imaging system and is equipped with a new type of static interferometric spectral payload, eliminating the moving parts of traditional spectrometers and significantly improving in orbit stability and reliability. Based on advanced Fourier transform interferometry technology, satellites can obtain continuous multi band information at once, achieving high-precision and high-efficiency acquisition of hyperspectral data. Its spectral performance is stable for a long time, requiring no frequent calibration in orbit, and can sustainably output high-quality space and spectral data. Relying on this new system, the star has a 10m hyperspectral spatial resolution in the visible near infrared spectrum, which is a major breakthrough in the domestic commercial aerospace hyperspectral remote sensing field and provides unprecedented data accuracy for quantitative remote sensing applications.
Secondly, in orbit intelligent computing: from "sensing the sky and calculating the earth" to "calculating the number of days".
The satellite is equipped with the Zhijia NX4 onboard intelligent computing unit, jointly developed with the Zhijiang Laboratory, with a computing power of 248 TOPS and an ultra high speed data interface. It can complete real-time preprocessing, target recognition, and information extraction of remote sensing data in orbit, achieving data acquisition, processing, and decision-making in the sky. It pushes remote sensing monitoring to a new stage of real-time active early warning, significantly shortening emergency response time and significantly reducing the pressure of satellite ground data transmission.
Thirdly, efficient global observation: from "single point coverage" to "global monitoring".
The star operates in a 535km sun synchronous orbit and can conduct high-frequency and regular monitoring of key areas. Single star is no longer limited to "single field applications", and can independently support the normalized observation needs of multiple fields such as ecological monitoring, ecological protection, urban governance, and agricultural and forestry monitoring in mining areas.
"One Soft" Empowerment: AI Models Activate the Value of Aerospace Data
If the "intelligent computing constellation" is the hard foundation for efficient data acquisition, then the self-developed "hyperspectral remote sensing AI model" is the soft engine that activates the value of data, solving the pain points of low utilization of massive hyperspectral data in the industry, relying on manual experience for interpretation, and difficult large-scale implementation.
China COSCO Solar Aerospace Joint Yangtze River Laboratory has built a hyperspectral remote sensing AI model, relying on its own high-quality and massive hyperspectral data, deeply integrating spectroscopy, artificial intelligence and industry knowledge, and training industry-specific hyperspectral AI models. In precision agriculture, ecological environment and other fields, it has achieved a leap from "observing growth" to "analyzing nutrition", from "checking pollution" to "identifying types", from "wide area search" to "targeted determination", making the spectral "fingerprint" more accurate, wider and better used, and connecting the front-end link of "on-board acquisition - in orbit interpretation". Satellites can independently output standardized research and judgment results.
Water monitoring is a typical landing scenario. It is reported that the Caiyun Hyperspectral 01 satellite is equipped with a jointly developed hyperspectral water sediment classification model, which adopts a lightweight in orbit inference scheme. It can directly distinguish the turbidity level of water bodies and mark suspected pollution areas in space. Only the classification conclusions and key area spectral data are transmitted, and fine depth analysis is carried out on the ground backend. This set of "in orbit pre-processing+ground precision processing" satellite ground collaborative mode is suitable for scenarios such as river control, flood control scheduling, and watershed ecological protection. The entire model supports in orbit iterative updates to continuously optimize recognition accuracy.
Against the backdrop of emphasizing the value of data in the 15th Five Year Plan, using massive and high-precision spectral data to train industry models, achieving a leap from "visible" to "understandable", this "soft engine" is becoming the core amplifier for aerospace data to move from "resources" to "assets".
Soft hardware collaboration: building a new ecosystem of "communication, computing, and remote"
Currently, commercial remote sensing is rapidly becoming popular worldwide, but traditional hyperspectral satellites generally suffer from pain points such as limited on-board storage, high transmission costs between satellite and ground, and long data processing cycles, making it difficult to meet the rapid response needs of emergency monitoring, precision agriculture, ecological environmental protection, and other scenarios.
Zhongkesi Aerospace has achieved precise breakthroughs through a dual line layout of "one hard and one soft": with "intelligent computing constellation" as the hard support and "hyperspectral AI model" as the soft engine, software and hardware collaboration and bidirectional empowerment have connected the entire chain of space and sky data collection and application. This layout not only meets the requirements of the development of new quality productivity, but also precisely solves the core contradiction of the industry's "massive data and lack of applications".
Qin Jing, Chairman and General Manager of China Aerospace Science and Technology Corporation, stated that in the face of the 15th Five Year Plan, China Aerospace Science and Technology Corporation has identified two main development lines: the "hard line" will continue to iterate intelligent hardware on board and strengthen real-time computing capabilities in space; The "soft mainline" iterates industry-specific hyperspectral models to deepen the landing of the remote sensing data industry. Zhongkesi Aerospace will work together with industry partners to continuously connect the integrated industry chain of "communication, remote sensing, and computing", and build a fully independent aerospace intelligent computing system that integrates "sensing, days, calculations, and applications". With domestically produced quantitative hyperspectral intelligent computing capabilities, it will empower various industries with digital monitoring and refined governance.
