Skip to content

Category: Uncategorized

Published on

What does mission-critical accuracy look like in practice?

Enabled Intelligence, an AI technology and services provider specialising in geospatial analysis, builds models for government and commercial clients across military intelligence, disaster recovery, transportation logistics and agricultural monitoring. Its quality bar is explicit: at least 95% accuracy on all training data, with certain ground truth datasets required to exceed 99%.

The technical constraints were as demanding as the accuracy targets. Datasets spanned synthetic aperture radar, electro-optical imagery, hyperspectral data and full motion video. Every project involved coordinate transitions: from geospatial coordinates for labeling, to pixel coordinates for model training, and back to geospatial coordinates at deployment.

After evaluating more than 35 labeling platforms and software packages, Enabled Intelligence selected Kili as the core of its annotation architecture. The team measures objects in real-world coordinates directly in the interface, stacks multiple image layers (synthetic aperture radar over electro-optical, for instance) and toggles between them during analysis, and runs quality control workflows tracking false positives, false negatives, misclassifications and label alignment.

The results: 95%+ accuracy across training data, 99%+ on ground truth datasets, and millions of labels produced across thousands of geospatial images for use cases ranging from aircraft detection to land use classification.

“We’ve assessed over 35 different platforms, labeling tools and software packages, and Kili has been by far the best platform that we have used for many of our types of data and certainly the best for geospatial,” says Peter Kant, CEO of Enabled Intelligence.

The full story is available in the Enabled Intelligence case study.

What have we shipped for EO teams since?

The lesson from programmes like this one is that geospatial data carries structure — spatial, geometric, radiometric — that generic image tooling ignores. Our recent releases follow that principle in four areas.

Geography as a way to organise the work, not just view it

A queue treats every asset the same way, in whatever order it happens to load. Geospatial datasets are not random: they have real spatial structure, and that structure is often exactly what should drive how work gets organised. Knowing that a cluster of assets sits over a flooded region, along a specific coastline, or across a city centre is the kind of context that should shape prioritisation and assignment, and it is precisely what a list view cannot surface.

Map View puts the dataset on an actual map. Teams see the geographic distribution at a glance and act on it directly, selecting, assigning or prioritising assets by where they are, with a built-in base map and reorderable layers underneath so the data reads correctly while they work.

Coordinates you can actually trust

Geospatial data is only useful if its coordinates hold up. Kili now preserves each image’s native coordinate reference system rather than forcing everything into a single projection, and teams can lock the CRS and resampling method at the project level so every asset is handled consistently. For imagery delivered with RPC metadata instead of full orthorectification, Kili supports an explicit affine transformation as a clearly flagged approximation, giving teams a usable path forward without quietly compromising on precision. The active coordinate system is always visible in the labeling interface.

Imagery rendered the way the data deserves

High bit-depth imagery (16-bit, Float32) carries far more dynamic range than standard 8-bit images. Kili now applies a percentile-based rescaling tuned for that range, so high bit-depth imagery renders with real contrast and detail, with no manual preprocessing required before import.

Confidence, visible right where labeling happens

Model predictions often come with a confidence score, but that signal is only useful if annotators can see it while they work. Kili now supports importing confidence scores and displaying them directly on geospatial objects in the labeling interface, giving teams a quick, built-in way to flag uncertain predictions for closer review.

Where this is heading

Across the EO ecosystem, the shift underway is from AI experiments to operational AI services, and it changes what dataset quality means. When a model informs a disaster response or a monitoring contract, training data becomes part of the supply chain: it needs provenance, consistent coordinate handling, and quality metrics that stand up to audit. The four capabilities above, like the workflows Enabled Intelligence built, exist to make that level of rigour the default rather than an achievement.

Kili Technology will keep developing along this line. EO organisations interested in the geospatial toolkit can find more detail on the geospatial annotation product page or get in touch with our team.

About Kili Technology

Kili Technology is the complete platform for building trustworthy, high-quality datasets for training, fine-tuning and evaluating AI/ML models, with dedicated support for geospatial imagery including SAR, electro-optical, hyperspectral and full motion video.

Published on

CloudFerro, a Polish sovereign cloud provider, has presented Bastion, a sovereign cloud architecture developed for defence, civil security organisations and operators of critical infrastructure.

Rather than being a standalone product, Bastion represents an operational cloud architecture designed for organisations that require full control over sensitive workloads, secure processing environments and resilient deployment models.

The solution combines cloud computing, AI capabilities and Earth Observation data processing within infrastructures operating under European jurisdiction. It can be deployed in different operational scenarios, ranging from permanent installations to isolated environments supporting mission-critical operations.

Earth Observation data increasingly plays a central role in civil protection, crisis response and security applications. Satellite imagery, together with AI-assisted analytics, enables faster situational awareness, infrastructure monitoring and decision support during emergencies.

Bastion builds on CloudFerro’s long-standing experience in delivering cloud infrastructure for European Earth Observation programmes and demonstrates how technologies originally developed for the EO sector can also support emerging dual-use applications while maintaining interoperability with broader European digital ecosystems.

The initiative reflects the growing importance of sovereign digital infrastructure as part of Europe’s resilience strategy, where secure cloud environments become an integral element of operational services based on satellite data.

More on: https://cloudferro.com/news/cloudferro-for-defence-and-national-security/ 

Published on

Monitoring protected nature is, at scale, mostly a screening problem. Municipalities are responsible for large areas, field visits are expensive, and working through maps and aerial photos by hand takes time. Changes are often caught late, or only when someone reports them.

VERNA does that screening automatically. Nature is notoriously complicated, and no single model handles all of it well. What sets VERNA apart is that it does not rely on one. It runs several deep-learning architectures side by side, for segmentation, change detection, embedding-based comparison and vegetation-height estimation, and turns what they find into results a caseworker can review and act on.

 

The inputs are aerial orthophotos and near-infrared imagery, airborne LiDAR and elevation models, and Sentinel imagery. The output is a shortlist of places where protected nature has probably changed: new water bodies, terrain that has been dug out or filled in, shifts in vegetation, habitats that are not registered yet.

 

That last step is the hard one. A prediction on a map is not useful by itself. A caseworker has to see why an area was flagged, compare it against earlier years, record an assessment and stand behind it later. VERNA is built around that whole process. In Denmark it has already identified more than 2,000 potential changes and 2,500 potential new ponds and lakes across 15 municipalities.

 

 

The goal is not to replace fieldwork. It is to make fieldwork more targeted. When authorities know where change is most likely, limited field resources can be used on the places where they matter most. That means earlier detection, better documentation and a more consistent basis for prioritising nature protection and restoration.

 

Denmark is a good place to build this because protected nature is already mapped in a structured way. The registrations of §3 nature, protected-area boundaries and habitat types give us unusually strong reference data to train models on and validate against. That is what deep learning needs to work well. It also lets us do something most monitoring cannot: cover large areas at high resolution and repeat it.

 

We have already taken VERNA beyond §3 nature into tree and canopy mapping. This works at the level of the individual tree: each tree is detected, located and measured. Combined with canopy-cover estimates, this makes 3-30-300 assessments more practical, because municipalities need both tree locations and canopy information to assess the rule consistently.

 

The next step is habitat screening for Annex IV species  – the strictly protected species that municipalities must account for in planning and casework. Here, the model does not replace field surveys. It predicts where suitable habitat is likely to occur, so a municipality knows where to look first.

 

There is a reason to do this now. The EU Nature Restoration Regulation has been in force since August 2024, and every member state has to submit a draft National Restoration Plan by September 2026. For urban ecosystems, the regulation also makes baselines unavoidable: Member States must ensure no net loss of urban green space and urban tree canopy cover by 2030, compared with 2024. All of it depends on a baseline you can defend and monitoring you can repeat over large areas.

 

Denmark is a good place to prove this, but the problem is not Danish. Across Europe, authorities are being asked to establish baselines, monitor change and prioritise restoration with limited capacity. The local rules, habitats and data differ, but the operational challenge is familiar: where should people look first?

 

That is where we think VERNA can travel. The approach is transferable, but it has to be adapted locally: finetuned on national data, mapped to local habitat definitions and validated with people who understand the landscape. We are looking for partners who can bring that local knowledge; authorities, EO companies, research groups and organisations working directly with nature monitoring.

 

About Koordinat

 

Koordinat is a Copenhagen-based AI and geospatial technology startup. We build tools that help authorities use Earth Observation data to monitor nature, identify changes and prioritise fieldwork.

 

Contact

 

Thomas Lykke Rasmussen

 

Co-founder, Koordinat

 

thomas@koordinat.ai

 

www.koordinat.ai <http://www.koordinat.ai/>

Published on

Within the framework of the West Africa Coastal Area Management Program (WACA), the CSE coordinates the West African Regional Coastal Observatory (WARCO). This regional observatory was established to strengthen coastal resilience by producing reliable, up-to-date, and accessible data, information, and knowledge across the entire West African coastal region. WARCO was established with financial support from the World Bank Group (WBG) and technical support from the West African Economic and Monetary Union (WAEMU) Commission. As part of WARCO’s mission implementation, the CSE selected VisioTerra for the oil spill detection and classification system in the Gulf of Guinea, named MONA-Oil. 

This choice follows all the studies carried out by VisioTerra to raise awareness of the numerous pollutants that threaten the entire ecosystem of the Gulf.

Delivered with a high-performance data center, MONA-Oil enables the acquisition of Sentinel-1/2 data for the Exclusive Economic Zones (EEZs) of the 12 countries of the West African coastal area as soon as it becomes available and processes it to detect dark objects that could be oil slicks.

AI algorithms are then used to classify these dark objects according to their potential origin: oil platform leak, tanker deballasting, natural oil seeps, or other.

If the dark object is correctly identified as an oil slick, a drift calculation module is applied, taking into account currents and winds.

An alert is then sent, especially if the oil slick could reach a coastline.

The MONA-Oil interface allows users to visualize these slicks as well as the areas that could be affected in the following days.

In all cases, validated detections are stored in a database to generate periodic bulletins and statistics by country, area, or type of pollution, for example. The delivery of the system and its data center is accompanied by training sessions in several of the 12 countries concerned: Benin, Gambia, Ghana, Guinea-Bissau, Republic of Guinea, Ivory Coast, Liberia, Mauritania, Sao Tome and Principe, Senegal, Sierra Leone and Togo.

MONA-Oil is the sixth system developed and delivered in Africa by VisioTerra, following FLEGT Watch for deforestation detection, MISBAR (*) for agriculture and irrigation, CAFWS (*) for forest management, GERNAC (*) for navigability in the Congo Basin, and TranshumAC for detecting transhumance routes and their intersections with cultivated or protected areas.

(*) These 3 systems were developed in partnership with the Moroccan company AfEOS.

Published on

In the high-stakes world of Space, we’ve spent decades playing a frustrating game of “wait and see” when it comes to Earth Observation (EO). You wait for the satellite to pass over, you wait for the clouds to clear, and you wait for the data to be processed into something actually useful. By the time you get your answer, the flood has usually already come and gone.

Enter RSS-Hydro. Based in the heart of Luxembourg’s booming space ecosystem, this geospatial analytics company isn't just processing data or running models; they are reimagining the very "recipe" of satellite intelligence. By 2026, their dual-threat strategy—the “multi-Sensor” and the “Pin” approach—is effectively flipping the EO market on its head.

Multi-Sensor Data: Moving Beyond the Single Ingredient

For years, the EO industry was siloed. You were either a "Radar person" (SAR) or an "Optical person." SAR could see through clouds but looked like grainy static to the untrained eye; Optical was beautiful and intuitive but useless the moment a storm rolled in.

RSS-Hydro’s multi-sensor or “data cocktail” approach treats these sensors not as rivals, but as ingredients. By using high-resolution SAR from the Copernicus Sentinel-1 missions as well as the rich spectral detail of Sentinel-2 with static hydrological layers from Digital Elevation Models (DEMs), and even satellite microwave data, they create a composite view that is greater than the sum of its parts.

This isn't just a simple overlay. Being able to deploy on various computing architectures, whether cloud or HPC supercomputers, RSS-Hydro employs a mix between signal processing and machine learning to fill in the gaps. The result? A clear, all-weather map that doesn't care if it's raining.

The ‘Pin’ Approach: Precision Where it Hits the Ground

If the multi-sensor is the what, the Pin approach is the where.

The traditional EO market sold "pixels" – large swaths of land without meaningful insights. But a city manager doesn't need a map of a province; they need to know if the water is going to hit the electrical substation on 5th Street.

The Pin approach "pins" satellite-derived (flood) intelligence to specific, high-value assets. It bridges the "last mile" of EO data by:

  • Localizing risk: Instead of broad area maps, it provides location-based (pinned) alerts for specific infrastructure using customizable parameters.
  • Integrating In-situ Sensors & Auxiliary Data: It can pull in data from IoT water-level sensors, from forecast simulations or even from citizen science-based collections to validate the satellite’s view in real-time.
  • Enabling LLMs to instantly interpret impacts: It converts large satellite imagery into lightweight, semantic "knowledge pins" either directly on a spacecraft or on the ground, enabling LLMs to instantly interpret impacts on specific assets or at specific locations through simple text-based data that can be processed on even the most resource-constrained platform or edge devices. 

"The EO market is shifting from selling images to selling impact. We don't just tell you it's flooding; we pin the risk to your front door." — Insight into the RSS-Hydro Philosophy.

Why the Market is Shifting

The commercial impact of these approaches will be profound. First responders, emergency managers, Insurance, transportation and supply logistics companies, all need to move away from static models toward the dynamic "Pin" for real-time alerting and planning. 

By breaking the "single-image" habit, RSS-Hydro has proved that the future of space isn't just about what we put into orbit - it’s about how we interpret the signals we’re already getting.

Published on

On 28th July of 2025 at night, a wildfire broke out near Mombeltrán (Ávila province, central Spain), rapidly spreading across the Sierra de Gredos and threatening nearby communities. OpenCosmos quickly activated its Hyper-500 product derived from hyperspectral Hammer satellite to capture imagery over the affected area. Within hours after the July 30th 15:13 UTC acquisition, the data was downloaded to the ground station, transferred, processed and analyzed, delineating the burn area and estimating the amount of hectares burned as of the image acquired during the event. 

Early detection provides a crucial window of opportunity for rapid response, allowing firefighting teams to attack the fire while it is still small and manageable. This proactive approach significantly increases the chances of successful containment, preventing the fire from escalating into an uncontrollable megafire. It drastically reduces the overall cost of suppression, minimizes damage to ecosystems and property, and most importantly, saves lives by enabling timely evacuations and ensuring the safety of first responders. 

The unique combination of the Hyper-500 constellation’s 32-band VNIR hyperspectral payload—delivering 5-meter resolution—and the OpenConstellation edge computing framework enables a paradigm shift in real-time intelligence. The satellites enable the deployment of algorithms directly to the spacecraft board, thus processing imagery on-board to generate low-latency insights. These 'lite' data products are then downlinked via Inter-Satellite Link (ISL) in a matter of minutes, effectively merging the company’s Earth Observation and telecommunications infrastructure into a single, high-speed backbone for rapid decision-making. 

Active wildfire fronts can be precisely identified and monitored using optical multispectral and hyperspectral satellite imagery. While Short-Wave Infrared (SWIR) and Thermal Infrared (TIR) bands are the primary domains for detecting thermal anomalies, hyperspectral data in the Visible and Near-Infrared (VNIR) range provides a unique chemical fingerprint of the combustion process. Specifically, it allows for the detection of narrow atomic emission peaks from elements like potassium (~766–770 nm), sodium (~819 nm), and calcium (~850–866 nm). Moreover, because longer wavelengths in the NIR and SWIR ranges penetrate smoke more effectively than visible light, these sensors can map the active flame front even when it is obscured by dense plumes. 

The acquired imagery is limited to the VNIR (Visible and Near-Infrared) spectrum. While True Color (RGB) composites failed to penetrate the dense smoke plumes generated by the active wildfire, the upper Red-Edge and NIR bands demonstrate a superior capacity for smoke penetration. In cases where smoke is not excessively thick, these longer wavelengths are less affected by Mie scattering, allowing for a clearer observation of the active fire front and underlying terrain. 

Figure 1: True Color Image in the affected area showing the smoke and False Color composite R, NIR, Blue 

Building upon these spectral signatures, an empirical detection method was developed by leveraging the Hyper-500’s 32-band VNIR range through advanced band algebra. By specifically isolating the channels corresponding to the potassium (766–770 nm) and calcium (~850 nm) emission peaks, the algorithm can automatically distinguish active hotspots from solar reflectance. This logic transforms hyperspectral data into a high-value 'lite' insight—a precise fire front delineation—which is then prioritized for immediate downlink via Inter-Satellite Link. This workflow ensures that critical fire-mapping intelligence reaches ground segments within minutes of acquisition, bypassing the latency of traditional raw data processing.

Figure 2: Normalized Differential NIR - Red Edge, and two close ups of several delineated wildfire fronts 

Some hyperspectral upper NIR and Red Edge bands have a peak of reflection that can differentiate the active fire front in comparison to other bands heavily influenced by chlorophyll scattering (Red Edge 2 and Red Edge 1) that doesn’t have the contribution of the peak of Potassium, Sodium or Calcium in the wildfire. 

These bands can be combined and operated through different analytics to delineate the active fronts and the burned area. 

Band Name Central wavelengthSpectral Range FWHM (Full Width at Half Maximum)Overlapping Chemical fingerprint
HS20 RedEdge 2 739nm 712,13 - 765,87 nm 26.87nm -
HS21 RedEdge 2 755nm 727,57 - 782,43 nm 27.43nmPotassium (~766–770 nm)
HS22 RedEdge 3 770nm 742,05 - 797,95 nm 27.95nm
HS23 RedEdge 3 785nm 756,52 - 813,48 28.48nm
HS24 Near Infra-Red 799nm 770,03 - 827,97 28.97nmPotassium (~766–770 nm) Sodium (~819 nm)
HS25 HS26 Near Infra-Red Near Infra-Red 814nm 830nm 787,51 - 843,49 nm 799,95 - 860,05 29.49nm 30.05nmSodium (~819 nm) 
HS27 Near Infra-Red 844nm 813,46 - 874,54 nm 30.54nmSodium (~819 nm) Calcium (~850–866 nm)
HS28 Near Infra-Red 860nm 828,9 - 891,10 nm 31.10nmCalcium (~850–866 nm) 
HS29 HS30 Near Infra-Red Near Infra-Red 874nm 884nm 842,41 - 905,59 nm 852,06 - 915,94 nm 31.59nm 31.94nm


The normalized differential index between the NIR band 30 and Red Edge band 20 in addition to a proxy derived NDVI (Band HS25 NIR and Band HS15 Red) to discard false positives, generates an SNR high enough to discriminate these NIR peak detections from healthy vegetation.

Figure 3: Normalized Differential NIR - Red Edge, False Color HS30, HS20, HS01 and detection over VHR pre-fire  condition  

Figure 4: a) Normalized Differential NIR (HS30) - Red Edge (HS20), False Color HS30, HS20, HS01 and detection over VHR pre-fire condition. 

A further investigation will be performed trying to correlate the peaks of reflectivity with the chemistry of combustion of Sodium, Calcium and Potassium and thus try to remove potential false positives and consolidate the approach. Moreover the technique will be refined lowering the threshold of detection and limiting the search only in close neighbour overlapping of 100-m buffer from the initial pixel detections.

The OpenCosmos Hyper-500 product offers high-resolution hyperspectral imagery, capturing 32 bands at under 5m resolution. The Technical Specifications of this constellation are detailed below: 

● 2 satellites 

● Ground Sampling Distance (GSD): 4.75m @ Nadir 

● Swath Width: 19 km 

● Bit Depth: 8, or 12 bits 

● Sun-Synchronous Orbit (SSO) at 14:00 hours Local Time Descending Node (LTDN). 

● Swath Resolution: 4096 pixels wide 

● Edge Computing with AI-onboard 

● IoT and Inter-Satellite-Link 

Spectral Bands 

The hyperspectral payload captures 32 channels in the visible and near infrared spectral range. The imager captures in linescan mode using a wedge filter. The 32 spectral bands central wavelengths can be chosen by the band start row, with the filter width being influenced by the number of dTDI stages in use. For simplicity, below the central wavelengths are listed at 1 dTDI stage with the nominal band start rows selected. 

● PAN: 625nm (FWHM: 250nm) 

● Band 0: 440nm - Blue (FWHM: 16.40nm) 

● Band 1: 455nm - Blue (FWHM: 16.93nm) 

● Band 2: 469nm - Blue (FWHM: 17.42nm) 

● Band 3: 485nm - Blue (FWHM: 17.98nm) 

● Band 4: 499nm - Blue (FWHM: 18.47nm) 

● Band 5: 515nm - Blue/Green (FWHM: 19.03nm) 

● Band 6: 529nm - Green (FWHM: 19.52nm) 

● Band 7: 545nm - Green (FWHM: 20.08nm) 

● Band 8: 560nm - Green (FWHM: 20.60nm)

● Band 9: 574nm - Green (FWHM: 21.09nm)

● Band 10: 589nm - Green/Yellow (FWHM: 21.62nm)

● Band 11: 605nm - Yellow (FWHM: 22.18nm)

● Band 12: 620nm - Yellow/Red (FWHM: 22.70nm)

● Band 13: 634nm - Red (FWHM: 23.19nm)

● Band 14: 649nm - Red (FWHM: 23.72nm)

● Band 15: 664nm - Red (FWHM: 24.24nm)

● Band 16: 679nm - Red (FWHM: 24.77nm)

● Band 17: 694nm - Red (FWHM: 25.29nm)

● Band 18: 709nm - RedEdge (FWHM: 25.82nm)

● Band 19: 724nm - RedEdge (FWHM: 26.34nm)

● Band 20: 739nm - RedEdge (FWHM: 26.87nm)

● Band 21: 755nm - RedEdge (FWHM: 27.43nm)

● Band 22: 770nm - RedEdge (FWHM: 27.95nm)

● Band 23: 785nm - RedEdge (FWHM: 28.48nm)

● Band 24: 799nm - Near Infra-Red (FWHM: 28.97nm)

● Band 25: 814nm - Near Infra-Red (FWHM: 29.49nm)

● Band 26: 830nm - Near Infra-Red (FWHM: 30.05nm)

● Band 27: 844nm - Near Infra-Red (FWHM: 30.54nm)

● Band 28: 860nm - Near Infra-Red (FWHM: 31.10nm)

● Band 29: 874nm - Near Infra-Red (FWHM: 31.59nm)

● Band 30: 884nm - Near Infra-Red (FWHM: 31.94nm)

Published on

Healthy benthic habitats are key to sustaining biodiversity, protecting coastlines, and  sustainable economic development. Accurate mapping of these benthic habitats is  essential for effective conservation and informed decision-making, yet conventional  mapping methods are often time-consuming and labor-intensive. 

Within the ESA Business Applications and Space Solutions (BASS) project SFC‑Online (Seafloor Classification Online), EOMAP – a Fugro company – addresses these challenges. By providing a cloud‑based software solution named ‘BENTHIQ’, EOMAP enables stakeholders to map and monitor benthic seafloor habitats efficiently and at  scale.  

The new service is designed to integrate Copernicus Sentinel‑2 data, very  high‑resolution commercial satellite imagery, and underwater video data into a  semi‑automatic workflow powered by machine learning.  

This workflow empowers users to generate high‑quality seafloor habitat maps without  requiring specialist expertise in Earth Observation or access to high‑performance local  computing infrastructure. ‘BENTHIQ’ is developed in close collaboration with the Norwegian Institute of Marine Research (IMR) and the State Office for the Environment  Schleswig-Holstein, Germany, who, in their role as pilot users, deliver training data and  support the development of the web application with their expert knowledge. 

The market opportunity for this online solution is substantial. By combining satellite with  underwater video data in a self-enabled, cloud-based platform, ‘BENTHIQ’ addresses an  unmet need for cost-efficient, scalable seafloor habitat monitoring. Leveraging EOMAP’s  established technological leadership and global reach, the service supports more  

frequent, reliable, and evidence-based marine management across governmental,  research, and industry users. 

Link Landing Page: 

https://benthiq.eoapp.de

Published on

The Greek national satellite space project: axis 3 land monitoring service aims to  strengthen the country's capabilities in satellite technologies and applications while  facilitating the exchange of satellite data. Its primary objective is to design, develop,  launch, and pre-operate small satellites capable of hosting multipurpose payloads to  address both national and European needs. 

The initiative seeks to provide high-resolution imagery to various Greek civil, institutional,  and governmental users, as well as potential European stakeholders within the  frameworks of Copernicus and GEOSS (Global Earth Observation System of Systems). 

Entirely funded by the European Union (EU) through the Recovery and Resilience Facility  (RRF), the project supports critical applications, including Land Use/Land Cover Mapping,  Deformation Monitoring, and Urban Analytics Services. 

------------------------------------------------------------------------------------------------------------- 

Consortis Geospatial (https://consortis-geo.gr/en) has played a pivotal role in advancing  geospatial technologies, particularly in change analysis. One of its key contributions is  the development of a change analysis algorithm, designed to enhance the accuracy and  efficiency of detecting and validating land surface transformations over time.

The main contribution of Consortis Geospatial was on land Cover Classification Change  Detection for Environmental and Resource Management. In that respect, Consortis  Geospatial led the: 

✔ Development of high-accuracy land cover maps to support sustainable environmental  monitoring and natural resource management. 

✔ Monitoring land use and land cover (LULC) changes to ensure long-term environmental  health and sustainability. 

✔ Development of a dedicated, tailored, software for LULC change detection, precision  assessment and validation with improved statistical metrics.  

Such change detection maps can contribute, to e.g., Agricultural Planning and  Sustainable Land Use through:

✔ Enhancing agricultural planning by distinguishing between vegetation types, optimizing  crop rotation, and improving soil management. 

✔ Promoting precision agriculture through detailed classification and mapping of  farmland. 

and to Disaster Risk Management and Early Warning Systems through: 

✔ Utilizing land cover classification and feature extraction for early warning systems in  detecting floods, wildfires, and other natural disasters. 

✔ Providing real-time geospatial insights to mitigate disaster risks, protecting both lives  and property. 

This advanced Land Cover change service by Consortis Geospatial will be a critical tool  for policymakers, land managers, urban planners, and researchers, enhancing land  management, disaster resilience, and environmental sustainability across Greece and  beyond. 

------------------------------------------------------------------------------------------------------------- 

“The project is being carried out under an ESA Contract in the frame of the Greek National  Satellite Space Project. The Project: Small-Satellites (Measure ID 16855) is implemented  by the Hellenic Ministry of Digital Governance with the European Space Agency (ESA)  Assistance in the Management and Implementation. The project is part of the National 

Recovery and Resilience Plan ‘Greece 2.0’, which is funded by the Recovery and  Resilience Facility (RRF), core programme of the European Union-NextGenerationEU. 

Views expressed herein can in no way be taken to reflect the official opinion of the  European Union/European Commission/European Space Agency/ Greek Ministry of  Digital Governance. Views and opinions expressed are those of the author(s) only and the  European Union/European Commission/European Space Agency/ Greek Ministry of  Digital Governance, cannot be held responsible for any use which may be made of the  information contained therein.”

Published on

GAF AG, an e-GEOS (Telespazio/ASI) company, has been awarded a World Bank-funded contract to modernise and expand Burkina Faso’s mining cadastre system eMC+ within the framework of the Support to Land and Mining Management Strengthening Project (PARGFM) at the Ministry of Economy, Finance and Development. The system update of GAF AG’s computerised mining cadastre platform eMC+ aims to reinforce efficiency, digital governance and customer service, thereby ensuring information transparency and accessibility within EITI-standards, and fostering land and mining cadastre interoperability in Burkina Faso.

The Ministry of Economy, Finance and Development of Burkina Faso’s “Support to Land and Mining Management Strengthening Project” (“Projet d’Appui au Reinforcement de la Gestion du Foncier et des Mines,” PARGFM) is being funded by the World Bank. PARGFM has contracted GAF AG to upgrade eMC+ + for the Mining Cadastre Department at the Ministry of Energy, Mines and Quarries to its latest version and implement additional modules of the cadastral platform, in order to ensure continuous operation and enhance transparency and the provision of rapid service by the Mining Cadastre General Directorate (DGCM).

The services being provided will be carried out in partnership with the Ministry of Mines and Quarries and under the supervision of its IT Services Directorate, within the framework of its IT master plan. They will include the development of new key features, such as electronic fee payments, improved customer notifications and interactive online services, with the aim of streamlining administrative procedures for mineral title applicants and holders in accordance with the latest requirements.

By accelerating the processing of mining titles and improving public access to information, the new eMC+ version is expected to further strengthen investor confidence and increase revenues for Burkina Faso’s mining sector.

The eMC+ system, which GAF AG first implemented in Burkina Faso in 2015, is a flagship product of GAF AG. It supports mineral tenure administration and public information access in line with EITI (Extractive Industries Transparency Initiative) standards. The system has been operational in Burkina Faso since 2018 and has a central role in managing and regulating the country’s mining titles.

The modernisation is intended to not only ensure continuous operation of the cadastre system but also to transfer technical skills to the Directorate-General of Mining Cadastre (DGCM), thus enabling full in-country management and sustainability within the context of the Ministry’s broader IT master plan.

The contract reinforces a decade-long consulting collaboration between the DGCM and GAF AG, marking a successful continuation of efforts to support a digital transformation and good governance for Burkina Faso’s mineral resources management.

Kick-off meeting at the Mining Cadastre office with the Secretary General of the Ministry of Energy, Mines, and Quarries, Mambagari COMBARI, the Director General of the Mining Cadastre, Mamadou SAGNON, and representatives of GAF AG.

About GAF AG

GAF AG, an e-GEOS S.p.A. (Telespazio S.p.A. /ASI) company based in Munich and Neustrelitz, was founded in Munich in 1985 as the first German company with a focus on applied remote sensing. It is one of the leading commercial geoinformation service providers in Europe. As part of the e-GEOS S.p.A./Telespazio S.p.A.  group of companies, GAF AG offers an extensive service portfolio that, in addition to the direct reception and distribution of satellite data, also includes highly developed analysis techniques, AI processes and the tailor-made development of geoinformation and software systems and platforms as well as comprehensive consulting solutions. The products and services in the sector of Advanced Air Mobility Solutions (drones) cover the entire value chain from data collection to system provision. The areas of thematic expertise for public and private clients worldwide include land monitoring, natural resource management, water and environmental monitoring, agriculture and forestry, mining, emergency management and infrastructure security. GAF AG is also one of the most experienced European service providers in the EU/ESA Copernicus programme, due to its many years of service implementation for the Copernicus land monitoring service, emergency management service and security and in-situ service components.

 

Published on

The European Space Agency will host StatEO26 – The EO for Official Statistics and Policy Indicators Reporting Conference at ESA-ESRIN on 5–7 May 2026. Co-organised with Eurostat (DG ESTAT), JRC, DG DEFIS, DG ENV, EARSC, EEA, OECD, UNSD, FAO, UNECE, the University of Hannover, and Biodiversity Alliance & CIAT, StatEO26 convenes national statistical offices, mapping and environmental agencies, EO providers, researchers and policy stakeholders to accelerate how satellite data feeds official statistics and policy reporting at national and international levels.

Why this conference matters

As governments scale up reporting on sustainability, environment and economy, Earth Observation (EO) offers consistent, timely and spatially rich evidence for indicators—supporting, for example, natural capital accounting, agricultural statistics, land-cover/use reporting, urban metrics, and GHG-related statistics. StatEO26 is designed to translate that promise into operational practice, focusing on methods, standards, uncertainty, and institutional uptake.

Call for abstracts (oral, poster & workshops)

The conference invites oral, poster, and workshop proposals. Submissions are especially welcome from teams demonstrating operational use in statistical production and from the Global South, highlighting capacity gaps and integration pathways with official systems. No special proceedings are foreseen. Submit via the conference portal. Key dates below.

Thematic oral sessions: Authors are encouraged to align with one of these tracks (each with a strong focus on methods, metadata and routes to official uptake):

  • Agriculture Statistics — crop type/area, yield, seasonal monitoring, change detection; integration with national crop tables and SDG/SEEA-related outputs.
  • Natural Capital Accounting — EO for SEEA ecosystem extent/condition/services; links to air, water, land and ocean accounts with traceable methods.
  • EO for SDGs & Environmental Policy Reporting — how EO supports national SDG and biodiversity reporting; processing chains and uncertainty handling.
  • Land Use / Land Cover (LULC) — operational classification & change detection; validation, INSPIRE links, reporting units (grid/NUTS), registries.
  • People & Urban Areas — population and settlement mapping, access to services, built-up area/green space per capita, heat-island, informal settlements.
  • Economy & Infrastructure — transport/industrial footprints, construction activity, night-time lights and other EO proxies for economic statistics.
  • Sustainability Indicators — forests, land degradation, biodiversity, emissions; validation, integration with national inventories and official outputs.

Interactive workshopsWorkshops are participatory (e.g., World Café/breakouts) and must deliver actionable recommendations. Propose one of the following:

  1. User needs & experiences (NSOs, mapping & environment agencies, etc.); success stories, gaps, procurement constraints.
  2. Integrating in-situ & EO — co-designed calibration/validation, governance, licensing and accessibility.
  3. Standardisation & quality — aligning EO with statistical quality frameworks, classifications and metadata.
  4. Trust & uncertainty — transparent methods, quality assessment, replicability and communication of uncertainty.
  5. Accessibility & interoperability — platforms, policies, metadata; Copernicus and other public EO services in data infrastructure.
  6. Capacity building — training and institutional development, with an emphasis on low-resource contexts.
  7. Future steps to operational integration — near-real-time EO, AI/digital twins, and sustaining publicly accessible EO datasets.

Who should contribute

  • National Statistical Offices, Mapping & Environment Agencies seeking scalable geospatial methods for official production.
  • EO companies and research groups delivering validated products and tools aligned with statistical standards.
  • Custodian agencies & international organisations working on comparable indicator frameworks and guidance.

Key dates

  • Call opens: 7 Oct 2025
  • Submission deadline (oral/poster/workshops): 1 Dec 2025
  • Notifications: 26 Jan 2026
  • Preliminary programme & registration open: 1 Feb 2026
  • Registration closes: 1 Apr 2026
  • Final programme: 7 Apr 2026
  • Conference: 5–7 May 2026 (Frascati, Italy)
  • StatEO26

Venue & contacts

ESA-ESRIN, Largo Galileo Galilei 1, 00044 Frascati, Italy.
Submissions & scientific queries: EO4Society.Conf@esa.int