AI-Powered Satellite PiSat-2 Embarks on Earth Observation Mission
The satellite began its journey on August 16 at 20:56 CEST (11:56 local time) aboard a SpaceX Falcon 9 rocket from Vandenberg Space Force Base in California, USA. Integrated by Exolaunch, PiSat-2 was part of the Transporter-11 rideshare mission, which also carried ESA's Arctic Weather Satellite.
At 21:50 CEST, PiSat-2 separated from the launch vehicle, and by 23:47 CEST, the Svalbard ground station in Norway confirmed it was safely in orbit, receiving a crucial signal from the satellite.
PiSat-2 is set to advance the field of Earth observation by showcasing how AI technologies can enhance the way we monitor our planet from space. The cubesat is equipped with a cutting-edge multispectral camera and a powerful AI computer capable of analyzing and processing images directly in orbit. This capability is expected to be vital for disaster response, maritime surveillance, environmental protection, and other critical applications.
"We are thrilled to launch PiSat-2, which will demonstrate the transformative power of artificial intelligence in Earth observation," said Simonetta Cheli, ESA's Director of Earth Observation Programmes. "This mission heralds a new era of actionable insights from space, promising smarter and more efficient ways of monitoring our planet."
Nicola Melega, ESA's PiSat-2 Technical Officer, added, "The PiSat-2 mission demonstrates how advanced AI technology can transform our ability to monitor and respond to changes on Earth, making space data more actionable and impactful than ever before."
A New Era for AI in Space
Artificial intelligence has greatly improved how we observe and understand our planet. AI enables satellite data to be processed swiftly and accurately, turning vast quantities of raw data into actionable insights for scientists, businesses, and policymakers.
While most AI processing occurs on Earth after data is downloaded, ESA's PiSat-2 will perform this processing in real-time in space. Instead of downlinking large volumes of raw data, including images obscured by clouds, PiSat-2 will process these images onboard, ensuring only essential information is transmitted to Earth. This innovation enhances data transmission efficiency and speeds up decision-making processes.
Orbiting at an altitude of 510 km, PiSat-2's multispectral camera captures images of Earth across seven bands, ranging from visible to near-infrared wavelengths. The 6U Cubesat platform, developed by Open Cosmos, operates AI applications that can be remotely installed and managed from Earth.
These applications provide actionable data for environmental monitoring and set a new benchmark in space-based AI technology. At launch, PiSat-2 includes the following applications:
- Cloud Detection: Unlike traditional satellites that transmit all captured images, PiSat-2 processes images in orbit, sending only clear, usable images to Earth. Developed by KP Labs, this application also classifies clouds and offers insights into cloud distribution, giving users flexibility in determining image usability.
- Street Map Generation: The Sat2Map application, created by CGI, converts satellite images into street maps. This feature is especially useful for emergency response teams, helping them identify accessible roads during disasters such as floods or earthquakes. The application will initially be demonstrated over Southeast Asia, showcasing its potential in crisis management.
- Maritime Vessel Detection: This application, developed by CEiiA, uses machine learning to detect and classify vessels in designated regions, supporting the monitoring of activities like illegal fishing. It underscores the satellite's role in promoting maritime security and environmental conservation.
- Onboard Image Compression and Reconstruction: GEO-K developed this application to compress images on board the satellite, significantly reducing file sizes and increasing data download speeds. After being transmitted to the ground, the images are reconstructed using a specialized decoder. The first demonstrations of this technology will focus on building detection in Europe.
To further enhance PiSat-2's capabilities, two additional applications will be uploaded now that the satellite is in orbit:
- Marine Anomaly Detection: Developed by IRT Saint Exupery Technical Research, this application uses machine learning to detect anomalies in marine ecosystems, identifying threats such as oil spills, harmful algae blooms, and heavy sediment discharges in real-time.
- Wildfire Detection: Created by Thales Alenia Space, this system uses machine learning to provide critical real-time information to firefighting teams. The tool generates a classification report to help locate wildfires, monitor fire spread, and identify potential hazards.
PiSat-2 is a collaborative effort led by Open Cosmos, supported by an industrial consortium that includes CGI, Simera, Ubotica, CEiiA, GEO-K, and KP-Labs.