CloudFerro and ESA push Earth observation forward with AI embeddings

In collaboration with the European Space Agency (ESA), CloudFerro has introduced a groundbreaking global AI embedding dataset for Earth observation (EO). By integrating cutting-edge AI technologies to enhance EO capabilities and provide more precise and scalable analysis of satellite data, global embeddings support EO-related research and application development. CloudFerro is a leading European cloud provider for the space sector. 

Embeddings are becoming increasingly significant in EO, offering a wide range of applications for professionals in the field. They can be used by remote sensing scientists, GIS analysts and environmental researchers working with satellite imagery and geospatial data.

“The traditional interaction of users with EO data is going to change dramatically with the wide introduction of embedding products at the scale of full Sentinel data archives. This prototype we built here is the first step towards understanding the value brought by this approach,” says Mikolaj Czerkawski from Φ-lab, an ESA research laboratory, who led the technical collaboration with CloudFerro. “By developing and releasing in a fully open-source setting, we demonstrate how the open data programmes like Copernicus, once again, can deliver unprecedented benefits to the wide community,” he continues.

Embeddings explained

Embeddings are high-dimensional vectors that transform raw data, such as images or documents, into numerical representations. This structured format captures relationships and semantic meaning within the data, allowing artificial intelligence (AI) models to process and analyse it with remarkable context awareness and precision. This enables machines to identify patterns, similarities and connections that might otherwise be challenging to detect.

This approach not only enhances the ability to process complex information, but also drives progress in areas such as machine learning, natural language understanding and computer vision. Embeddings provide the foundation for scalable, versatile AI solutions, unlocking new possibilities across a wide range of applications, from predictive modelling to advanced decision-making systems.

“The Sentinel data embeddings generated with Major TOM and hosted on our CREODIAS platform will bring new capabilities to the geospatial community by making high-quality, AI-ready data accessible globally,” states Dr Jędrzej Bojanowski, data science manager from CloudFerro. “This collaboration with ESA highlights our commitment to embrace the AI revolution and introduce it to EO data ecosystem at large, including Copernicus,” adds the expert.

“With this release, ESA is adding more momentum to these efforts to help secure a strong position of the European entities in this area,” comments Anna Burzykowska, Copernicus innovation officer at ESA. “We are keen to continue our collaboration with our industrial and research partners and work diligently to lay the key foundation needed to grow the core of this technology here in Europe, especially for the Copernicus programme.”

Powered by technology

The project leverages GPU-accelerated cloud infrastructure provided by CloudFerro on the CREODIAS platform, enabling fast and efficient processing of massive datasets. In collaboration with ESA’s Φ-lab, advanced computer vision models such as DINOv2, SigLIP and SSL4EO were implemented, forming the foundation of AI solutions in EO.

Over 62TB of raw satellite data from the Copernicus programme were processed in the project, generating more than 170 million embeddings. This process condensed 9.368 trillion pixels into 1TB of optimized data, which is now easy to analyse and integrate into AI models. In the next phases of the project, the team will test the performance of embeddings in tasks such as pattern detection and predictive modelling.

Click here for more information on the global embeddings dataset for Earth observation.