European Space Agency (ESA) has launched an AI powered satellite for Earth observation. The CubeSat embarked on its journey into space on 16 August at 20:56 CEST on board a SpaceX Falcon 9 rocket from the Vandenberg Space Force Base in California, U.S., integrated by Exolaunch as part of the Transporter-11 rideshare mission, which also included ESA's Arctic Weather Satellite.

The agency believes that the Φsat-2 will push the boundaries of Earth observation, demonstrating how AI technologies can be used to advance how we observe Earth from space. The miniature satellite has a state-of-the-art multispectral camera and a powerful AI computer that analyzes and processes imagery while in orbit. This will prove critical for disaster response efforts, maritime monitoring, environmental protection and more.

According to ESA's Director of Earth Observation Programs, Simonetta Cheli, they are thrilled to launch Φsat-2, which will demonstrate the transformative power of artificial intelligence in Earth observation. In their opinion, this mission heralds a new era of actionable insights from space, promising smarter and more efficient monitoring of our planet.

Era of AI

AI has significantly enhanced the way our planet is monitored and understood. AI capabilities allow satellite data to be processed quickly and accurately, transforming vast amounts of raw data into actionable insights for scientists, businesses, and policymakers.

While most AI processing occurs on the ground after downloading the data, ESA's Φsat-2 satellite will do this in real-time. Instead of downlinking large amounts of raw data, including those obscured by clouds, the onboard apps can process these images directly, ensuring that only the most essential information is sent back to Earth. This innovation gives tangible benefits, enhances data transmission and accelerates decision-making.

The satellite orbits Earth at an altitude of 510 km. Its multispectral camera images Earth in seven different bands in the visible to near-infrared part of the electromagnetic spectrum. OpenCosmos designed and developed the 6U CubeSat platform, which runs AI apps that can be easily installed and operated remotely from Earth.

These apps can provide actionable data for environmental protection and set a new standard in space-based AI technology. Unlike traditional satellites that downlink all captured images, including those obscured by clouds, Φsat-2 processes these images directly in orbit, ensuring that only clear, usable images are sent back to Earth. The Sat2Map application, developed by CGI, converts satellite imagery into street maps. The maritime vessel detection application, developed by CEiiA, utilizes machine learning techniques to automatically detect and classify vessels in specified regions, facilitating monitoring activities like illegal fishing. Onboard image compression and reconstruction, developed by GEO-Kis responsible for compressing images on board. By significantly reducing file sizes, this application increases the volume and speed of data downloads. After being downlinked to the ground, the images are reconstructed using a dedicated decoder. 

Phys Org