Europe's First Artificial Intelligence Space Missions

With Europe's first AI Earth observation mission ɸ-sat-1 in space, ɸ-sat-2 is now underway.
Susan Fourtané
Europe's ɸ-sat-1 launch onboard Vega rocketESA-European Space Agency 

Europe's first Artificial Intelligence Earth observation mission, ɸ-Sat-1, was successfully launched. ESA's ɸ-Sat-2 is now underway. 

Following ESA's successful launch of its first Artificial Intelligence (AI) Earth observation mission ɸ-Sat-1 (pronounced Phi-Sat-1), the European Space Agency (ESA) follows-up with the next innovative state-of-the-art technology: ɸ-Sat-2 (Phi-Sat-2).

In another historical event to advance European space missions, on September 3, 2020, ESA launched its ɸ-Sat-1, an enhancement of the Federated Satellite Systems mission on-board a Vega rocket from Europe's spaceport in Kourou, French Guiana. 

ɸ-Sat-1 mission is to demonstrate how satellite data, coupled with advanced onboard digital technologies, can bring benefits to businesses, industry, and science.

Europe's first Artificial Intelligence in space's mission: AI cloud detection, atmospheric and space science observations

Europe's First Artificial Intelligence Space Missions
PICosatellite for Atmospheric and Space Science Observations (PICASSO CubeSat) designed to investigate the upper layers of Earth's atmosphere, Source: ©BISA/ESA

The main task of the AI chip on ɸ-Sat-1 is to comb through huge sets of images and filter out the ones of low-quality due to cloud coverage.

The main advantage of processing large amounts of data on-board is that it makes the delivery to Earth more efficient as the AI has already removed the cloudy images. 

The AI cloud detection experiment is going to validate the performance of the on-board inference engine based on a machine learning algorithm for cloud detection.

ɸ-Sat-1 AI

Europe's First Artificial Intelligence Space Missions
This is the small yet powerful ɸ-Sat-1 AI, in case you were fantasizing with a futuristic AI robot, Source: CERN/M. Brice, ESA

 ESA has been working with various partners to demonstrate the potential of Artificial Intelligence and its applications in space to develop ɸ-Sat and to enhance the FSSCat mission. 

The hyperspectral camera on one of the two CubeSats is going to collect a huge number of images of planet Earth. Some of these images will have to be discarded due to cloud coverage. 

As a way of saving time and human effort back on Earth, the ɸ-Sat Artificial Intelligence chip is going to identify the unsuitable images and filter them out, only sending the images which contain usable data. 

"The great interest and learning experience we gathered with ɸ-Sat-1 encouraged us to continue with ɸ-Sat-2. The positive feedback we received after evaluating 16 great mission concepts for ɸ-Sat-2 gives us the signal to continue preparing AI technology and issuing more ɸ-Sat calls in the coming years, said Josep Rosello, Head of Technology Coordination and Frequency Management Section at ESA.

"ɸ-Sat-2 is the next step in ESA's drive for continuous innovation in Earth observation, which was initiated with the creation of the ɸ-Lab in 2017, and the launch of ɸ-Sat-1 on September 3 this year. ɸ-Sat-2 will further push the frontier of technology and open up new opportunities for the space and data analytics industry," added ESA's Director for Earth Observation, Josef Aschbacher.

ɸ-Sat-2 AI

The ɸ-Sat-2 mission will further demonstrate the capabilities of Artificial Intelligence (AI) technology for Earth observations following up from ɸ-Sat-1. The use of AI is going to lead to novel ways of collecting, distributing, and analyzing data collected in space about planet Earth. 

Now ESA has announced that a panel of ESA experts has selected the mission proposed by a consortium as the winning idea. The consortium is led by Open Cosmos. 

The consortium includes six different European countries and includes CGI, Ubotica, Simera CH Innovative, CEiiA, GEO-K, and KP Labs as partners and sub-contractors. If negotiations are successful, the ɸ-sat-2 AI satellite will be ready to start its mission just 16 months after the agreement.

The new ɸ-Sat-2 mission has been planned to address a wide range of Artificial Intelligence applications which include transforming a satellite image into a street map, cloud detection in order to reduce the huge volume of data that needs to be downloaded to the ground base, autonomous detection and classification of maritime vessels, forest monitoring and anomaly detection.

Flying in a Sun-synchronous orbit, the payload will include the AI processor Intel Movidius Myriad 2 from Ubotica. The AI processor was already adopted on the ɸ-Sat-1 mission.

Expressing his excitement about the successful launch of the ɸ-Sat-1, Massimiliano Pastera, ɸ-Sat-1 and ɸ-Sat-2 Officer at ESA said that the launch of ɸ-Sat-1 will allow them "to understand the use of AI for cloud detection, and ɸ-Sat-2 will represent a flying platform and give us the opportunity to experiment with multiple applications, as well as verify the enabling capabilities of on-board AI for Earth observation."

Related Articles: 

Add Interesting Engineering to your Google News feed.
Add Interesting Engineering to your Google News feed.
message circleSHOW COMMENT (1)chevron
Job Board