China Launches First Quantitative Hyperspectral Satellite with On-Orbit AI Computing
China successfully launched the "Xiguang-2 01" (Caiyun Hyperspectral 01) satellite, its first quantitative hyperspectral satellite featuring integrated AI computing for real-time, on-orbit data processing. Developed by Zhongke Xiguang Aerospace, the satellite was launched on July 22, 2026, using a Gravity-1 Yao-4 carrier rocket. This launch marks a significant advancement in edge computing and remote sensing, enabling faster data processing directly in space instead of relying solely on ground stations. It enhances China's capabilities in high-resolution commercial Earth observation and resource exploration. The satellite achieves a 10-meter spatial resolution and a 7-nanometer spectral resolution, capturing up to 1 million square kilometers of imagery daily. It utilizes a new interference spectral imaging technology that allows the center wavelength to be freely defined for precise remote sensing applications.
## BACKGROUND
Hyperspectral remote sensing collects and processes information from across the electromagnetic spectrum, allowing for the identification of materials and minerals that are invisible to the human eye. Traditionally, satellites transmit raw data to Earth for processing, but integrating on-orbit AI computing (edge AI) helps bypass downlink bandwidth bottlenecks by processing data directly in space.