ExtremeEarth · Project

Satellite Big Data Analytics Turning Petabytes of Earth Images Into Actionable Intelligence

digitalTestedTRL 5

Imagine Europe has satellites photographing the entire planet every few days — that creates petabytes of images, but most of it just sits there because it's too massive to analyze. ExtremeEarth built AI-powered software that can automatically sift through all that satellite data and pull out useful information, like crop health maps or ice sheet changes. Think of it as a super-powered search engine, but instead of searching the web, it searches satellite pictures of Earth. The team demonstrated this with two real use cases: monitoring food security and tracking polar ice changes.

By the numbers

Petabytes

Volume of Copernicus satellite data the platform is designed to process

Consortium partners across the project

Countries represented in the consortium

Demonstrated use cases (Food Security and Polar)

Total project deliverables produced

Demo-type deliverables with working software

Industrial partners committed to commercial exploitation

The business problem

What needed solving

Companies in agriculture, insurance, and environmental services increasingly depend on satellite imagery for decisions — but Copernicus satellites generate petabytes of data that overwhelm conventional analysis tools. Most organizations lack the AI and big data infrastructure to extract timely, actionable insights from this flood of Earth observation data. The result: expensive satellite data sits unused, and decisions get made on incomplete or outdated information.

The solution

What was built

The project built a complete big data analytics platform for Earth observation: Strabon for querying massive geospatial datasets, JedAI for interlinking data sources, SemaGrow for federating distributed data, GeoTriples for converting geospatial data to linked formats, a Hops data platform integrating all components, a semantic catalogue for discovery, deep learning models with a large training database, and an evaluation benchmarking system (Geographica). All core tools were delivered in two versions with the final versions evaluated at scale.

Audience

Who needs this

Precision agriculture companies monitoring crop health at continental scaleReinsurance firms modeling climate and polar ice riskEnvironmental consultancies doing satellite-based land monitoringGovernment agencies managing food security programsEarth observation data service providers looking for better analytics backends

Business applications

Who can put this to work

Agriculture & Crop Monitoring

mid-size

Target: Precision agriculture companies and agri-food corporations managing large-scale crop portfolios

If you are an agri-tech company struggling to monitor crop health across thousands of hectares — this project developed deep learning models and a large training database specifically for the Food Security use case, turning raw Copernicus satellite images into actionable crop intelligence. The platform processes petabytes of Earth observation data automatically, replacing manual image analysis. With 11 consortium partners across 6 countries validating the approach, this is field-tested technology.

Insurance & Climate Risk

enterprise

Target: Reinsurance firms and catastrophe modelers assessing environmental exposure

If you are an insurance company trying to price climate risk but drowning in satellite data you cannot process fast enough — ExtremeEarth built a distributed analytics platform (Strabon, SemaGrow, Hops) that queries and federates massive geospatial datasets at scale. The Polar use case demonstrated tracking ice sheet changes, directly relevant to sea-level rise modeling. The semantic catalogue lets you search across linked geospatial data sources instead of working with disconnected datasets.

Environmental Consulting & Monitoring

any

Target: Environmental services firms contracted for land use monitoring, deforestation tracking, or regulatory compliance

If you are an environmental consultancy spending weeks manually analyzing satellite imagery for client reports — this project delivered 29 deliverables including software for querying big linked geospatial data, interlinking data sources (JedAI), and transforming geospatial data into linked formats (GeoTriples). The 2 demonstrated use cases — Food Security and Polar — show the platform works on real-world environmental monitoring tasks, not just in the lab.

Frequently asked

Quick answers

What would it cost to adopt this technology?

The project was a Research and Innovation Action (RIA), meaning the outputs are primarily research software tools. Licensing terms would need to be negotiated directly with the consortium, led by the National and Kapodistrian University of Athens. Since 3 industrial partners were involved, some components may already have commercial licensing paths.

Can this handle industrial-scale satellite data volumes?

Yes — the project was specifically designed for petabytes of Copernicus data, which is about as large-scale as Earth observation gets. The distributed implementations of Strabon (querying) and SemaGrow (federation) were built in two versions, with the final versions evaluated for performance at scale. The Hops data platform was developed specifically to integrate all components for production-like workloads.

What about IP and licensing — who owns this technology?

As an EU-funded RIA project, the IP typically stays with the consortium partners who developed each component. With 4 universities and 4 research organizations in the consortium, expect a mix of open-source tools and proprietary components. The 3 industrial partners were explicitly tasked with commercial exploitation of results.

How mature is this — is it ready to deploy or still experimental?

The project delivered final versions (version II) of all major software components and demonstrated them in 2 complete use cases (Food Security and Polar). This puts it past pure research but likely still needs integration work for a specific commercial deployment. Based on available project data, expect a customization and integration phase before production use.

Does this work with existing geospatial data infrastructure?

The platform was built to work with Copernicus data on existing infrastructure including Polar TEP and DIAS (Copernicus Data and Information Access Services). GeoTriples converts geospatial data into RDF linked data format, and SemaGrow federates across multiple data sources, so it is designed to connect with — not replace — your existing data systems.

What specific AI capabilities does this provide?

The project built deep learning models trained on a large training database specifically for Earth observation tasks. These models extract information from satellite imagery for the Food Security use case (crop monitoring) and the Polar use case (ice monitoring). The AI components are integrated into the broader analytics platform rather than being standalone tools.

Consortium

Who built it

The ExtremeEarth consortium brings together 11 partners from 6 countries (Germany, Greece, Italy, Norway, Sweden, UK), with a balanced mix of 4 universities, 4 research organizations, and 3 industrial partners — giving a 27% industry ratio. The project is coordinated by the National and Kapodistrian University of Athens, a major Greek research university. The 2 SMEs in the consortium suggest some startup-level commercial interest, though the project leans academic. For a business buyer, the key question is which of the 3 industrial partners holds the commercial rights to specific components and is positioned to offer them as a service or product.

ETHNIKO KAI KAPODISTRIAKO PANEPISTIMIO ATHINONCoordinator · EL
UNIVERSITETET I TROMSOE - NORGES ARKTISKE UNIVERSITETparticipant · NO
KUNGLIGA TEKNISKA HOEGSKOLANparticipant · SE
HOPSWORKS ABparticipant · SE
DEUTSCHES ZENTRUM FUR LUFT - UND RAUMFAHRT EVparticipant · DE
UNITED KINGDOM RESEARCH AND INNOVATIONparticipant · UK
NATIONAL CENTER FOR SCIENTIFIC RESEARCH "DEMOKRITOS"participant · EL
UNIVERSITA DEGLI STUDI DI TRENTOparticipant · IT
METEOROLOGISK INSTITUTTparticipant · NO
VISTA GEOWISSENSCHAFTLICHE FERNERKUNDUNG GMBHparticipant · DE

How to reach the team

The coordinator is the National and Kapodistrian University of Athens (Greece). SciTransfer can facilitate an introduction to the right team member.

Order a report on this consortium See ETHNIKO KAI KAPODISTRIAKO PANEPISTIMIO ATHINON profile →

Next steps

Talk to the team behind this work.

Want to explore how ExtremeEarth's satellite analytics could work for your business? SciTransfer can connect you directly with the project team and help assess fit for your specific use case.

Order a report on this topic More in Digital & ICT