Archives: Activities Portfolio

RESTART

Objectives of the Product

Cotton farm professionals encounter interconnected agronomic and financial challenges, including unpredictable crop health, pest infestations, water shortages, and fluctuating market prices. RESTART aims to directly tackle these challenges by providing:

  • Crop Monitoring, Pest and Irrigation Alerts: by utilising Earth Observation (EO) and meteorological data, farmers obtain timely notifications regarding crop stress, pest development, and irrigation needs, enhancing field management practices.
  • Yield Prediction: in-season AI models combine remote sensing and agronomic data to produce precise, location-specific yield projections.
  • Financial Forecasting: analytics of the cotton market, along with predictive modeling and hedging strategies, enable farmers to optimise the timing of sales and safeguard profit margins.

RESTART inherits knowledge from previous ESA and EU-funded projects, effectively combining agronomic intelligence and financial forecasting within a scalable, tiered-subscription platform. The system is designed for wide accessibility, suiting both smallholders and large cooperatives.

Customers and their Needs

RESTART targets smallholder cotton farmers, agricultural cooperatives and organisations and large agribusinesses. In cotton farming, it is common for one entity to manage everything from budgeting and input scheduling to production planning. These raise the need for practical, holistic insights—such as crop health alerts, irrigation needs, pest outbreaks predictions, and yield forecasts—to enhance productivity and reduce costs. The financial forecasting service provided by RESTART is essential, as customers usually lack knowledge in economics, trading and hedging strategies, leading to poor decisions regarding the crop market. In addition, individuals possess inadequate expertise in satellite and/or AI technology, emphasising the need for user-friendly and comprehensive dashboards and notifications. Consultants and traders will benefit from RESTART, as their roles need insightful, scalable tools to assist multiple farms or large agribusinesses.

Targeted customer/users countries

Greece, Europe, scalable for global coverage.

Product description

RESTART is an AI-driven decision support platform based on EO satellite imagery, real-time meteorological and market data. Its services include:

  1. Crop Monitoring
  2. Predictive Alerts for pest outbreaks
  3. Predictive Alerts for water stress
  4. Yield Forecasting using seasonally adaptive AI models
  5. Financial Insights through cotton futures price forecasting and market insights.

RESTART stands as a first-of-its-kind platform due to its:

  • Holistic cotton farm management, including the consolidation of agronomic and financial forecasting within a single platform.
  • Advanced multi-parametric modelling, integrating a wide range of input data including EO, pest scouting, weather, historical yield data, soil moisture levels, financial markets’ data and other economic indicators.
  • Proactive modelling, anticipating risk before onset.
  • Scalable cloud architecture that accommodates both smallholders and large farms.
  • User-centric interface that provides insights through web dashboards, alerts, and APIs.

Farmers engage via a platform that provides insightful dashboards, interactive visualisations, and alert configurations. The platform’s architecture consists of:

  1. Data ingestion layer encompassing satellite, weather, and market APIs
  2. Agronomic and financial processing modules (model training and forecasting)
  3. Storage backend for time-series and historical data
  4. Web and API services, including dashboards, visualisations, and alerting functionalities.
Added Value

RESTART offers distinctive added value compared to existing AgriTech platforms through the following key differentiators:

1. Integrated Agronomic and Financial Intelligence into a Unified Platform

Unlike other platforms that focus exclusively on agronomic or financial advisory systems, RESTART uniquely combines satellite-derived agronomic risk models with AI-driven cotton price forecasting. This integration enables farmers to align field operations with financial decision-making in order to optimise cotton farm management.

2. Predictive Analytics Across Diverse Data Sources

RESTART utilises EO, weather, and financial datasets to improve essential agricultural and business operations, including crop monitoring, yield prediction, pest and irrigation alerts and cotton price forecasting. This AI-driven, multi-dimensional method surpasses single-focus tools by providing proactive, accurate insights that help professionals optimise both field performance and business outcomes.

3. Scalable and Accessible Service

RESTART provides scalable, cloud-based analytics through user-friendly web visualisations and automated alerts designed for both smallholders and large farms, enhancing decision-making promptness and adoption rates.

 4. Cost-Effective, Tiered Subscription Framework

RESTART offers a flexible pricing structure tailored for different cotton farm sizes, thereby democratising access to advanced analytics. This approach ensures affordability in contrast to high-cost competitor enterprise solutions, while maintaining high service value and return on investment.

Current Status

The RESTART activity is currently in its De-risking Cycle (M1–M10).

Ellipsis Map Engine

Objectives of the Product

Existing tabular compute engines are not built for raster data.

The problem is that existing compute engines force users to manually catalogue, shard, and mosaic their raster data for distributed compute and integration. This slows down spatial data science, making professionals lose weeks to manual data preparation and awkward workarounds, instead of gaining insights and scaling capabilities.

As a solution, Ellipsis Map Engine brings users map-native, distributed analysis to all their geospatial data types so they can run their workflows instantly, without the manual work. It’s the missing piece, the fit-for-purpose raster data Lakehouse that brings users fast and easy spatial data science at scale:

  • Automate raster data ingestion, management and distribution
  • Run scalable, flexible and interactive spatial analytics effortlessly
  • Get insights across GIS & non-GIS systems instantly
Customers and their Needs

The targeted user segments are Space/EO-powered value-adders, Civil Engineering – environmental (hazard), water and/or geotechnical engineering and consulting -, and P&C/Ag. insurance (insurance, reinsurance, GIS/NatCat data and/or consulting services). 

Because the state of the art in performant geospatial data analytics currently lacks the possibility to create raster-native frameworks, the existing table-native engines are unable to automatically ingest and spatially shard raster data, leaving the job of catalogue building, sharding and mosaicking to the end user. This is a major handicap for our target customer segments (all professionals who rely on EO/spatial data and who want to use and integrate this into data science workflows, either for internal use or to provide better/new services to clients). The lack of infrastructure that supports such interactive data science on raster data is increasingly painful because EO and environmental data (commonly found in raster format) are increasingly relevant due to climate change and environmental risks that are affecting the industries of our target customers.

End users (representing each of the target segments) are involved in the activity as pilot users and will provide us with feedback and validation throughout the project.

Targeted customer/users countries

Committed pilot users are confirmed for The Netherlands and France. Our target segments have a global presence and Ellipsis Map Engine, as an infrastructure solution, is not geographically limited.

Product description

Under this activity, we are building the Ellipsis Map Engine. This product complements existing table engines by uniquely allowing users to load any EO and spatial data into a cluster, and have each node automatically host a section of the dataset based on geography (spatial sharding). So, when a Python command is created, it can be run both rapidly and geospatially aware, as each node in the cluster only needs to execute the command for the geographic section it has loaded. This simple but effective strategy supports use cases in which ever changing spatial logic needs to be applied to large raster/spatial data on-the-fly.

Our team’s existing product, Ellipsis Drive, is a central repository for discovering, managing and consuming spatial data across teams, organisations and workflows. Clients pay a monthly/quarterly/yearly fee to use Ellipsis Drive to search, access and leverage their EO/spatial datasets as high performance and interoperable web services (storage-based PaaS model). The Ellipsis Map Engine is an extension to this existing solution, adding the option to apply flexible and high-performance analytics to the spatial data that customers are hosting (adding a fee for use of Processing Units under the same Paas model).

Added Value

Users can automatically run arbitrary logic on any EO/spatial data in a distributed and spatially aware way and use results in both GIS and non-GIS systems instantaneously. This unique capability allows them to get answers to ad hoc spatial analysis challenges easily and time efficiently, just like they are used to when analysing table-native data with regular tabular engines. Thus, this tool creates value for Data Scientists, Actuaries and Geo Engineers who work with EO and EO-derived (raster) data, allowing these professionals to apply complex – and ever changing – analytics to spatial/EO data in a highly time-efficient way.

For Data Scientists, consultants, and actuaries in P&C/Ag. (re)insurance, this directly supports highly effective pricing, (aggregate) risk mapping, portfolio management and claims operations (all processes relying on EO/raster data and ad hoc spatial logic).

For Data Scientists, Consultants and (Geo)Engineers in civil engineering this means getting the answers to their client’s questions much more quickly and accurately, including for governments, environmental agencies as well as energy and utility companies. For Data Scientists and Consultants working at EO value-adding companies, providing flexible and ad hoc EO-powered data services is much easier and more cost efficient.

Current Status

The activity has just kicked off. The project team is working to deliver on its first milestone, currently conducting market research for requirement definition and prioritisation, creating and fine-tuning the pricing model, and building fitting B2B go-to-market alliances and channel partnerships. Soon we will also start with technical storage and compute design, product development and the creation of detailed user stories.

FANTOM

Objectives of the Product

FANTOM supports the Department for Environment, Food & Rural Affairs ambition “to make our air purer, our water cleaner, our land greener and our food more sustainable. Our mission is to restore and enhance the environment for the next generation, and to leave the environment in a better state than we found it.”

FANTOM builds a high spatial and temporal resolution UK-centric database of agriculture and biodiversity markers with deep thematic content and context to support schemes aimed at the UK’s Agricultural Transition as well as supporting Net Zero and climate change activities.

These markers, schemes and impact assessment layers will be available to all government agencies, associated arms-length bodies, and commercial companies to be able to monitor and measure the progress of their sustainability activities and interventions.

Initially focused on delivering markers to support schemes for agricultural Paying Agencies, the aim is to ensure that the markers and schemes have a wider commercial need across government, non-government and commercial markets.

Furthermore, it will change the current perception of Earth observation-based land monitoring that not only looks at the asset but also its relationship with neighbouring assets to support landscape-level sustainability activities across the wider landowning community.

Customers and their Needs

FANTOM’s primary intended customers are agricultural Paying Agencies in the United Kingdom and the European Union.

First, these organisations deal with a huge number of landowners and require ongoing information on their customer base. This assessment demands automated and reliable satellite data analytics to ensure that subsidies and interventions are made correctly and can be monitored.

Second, there are a wide range of users interested in monitoring their land to evaluate its habitat quality and biodiversity value. This includes Paying Agencies but also other government bodies such as Environmental Ministries plus commercial actors such as developers, infrastructure owners and more. These users face the challenge of understanding the impact of their actions on the wider landscape.

Targeted customer/users countries

UK, EU

Product description

FANTOM delivers high spatial and temporal resolution markers to support agri-environmental schemes through the analysis of EO data and its integration with ancillary information.

The main subsystems and components of FANTOM are:

1. Data Ingestion and storage: handled primarily by Earth-i’s existing satellite and in situ data processing workflow with necessary customisation for FANTOM.

2. Data Processing:

Depending on the marker requirements:

  • Processing scripts capture the workflow and procedures to be applied to the data to realise the markers and evaluations;
  • An AI toolset provides the facility to analyse imagery using tools such as computer vision, machine learning and neural networks;
  • Scheme scripts will combine information about multiple markers to perform scheme assessments.

3. Data Analysis: analytics performed using a combination of sophisticated computer vision techniques and machine learning models to pick out specific identifiable features in the source satellite data. Outputs the markers required for the individual scheme assessments.

4. QA/QC Pipelines: QA/QC pipelines allow both manual and automated intervention.

5. Data visualisation: using existing platforms and open-source tools, visualisation will allow the user to overlay the resulting markers on the map, chart the outcomes and make rapid statistical assessments of marker outcomes and scheme results.

Added Value

FANTOM will reduce the amount of time that it would otherwise take Paying Agencies to generate a full and comprehensive set of monitoring indicators for the new Environmental Land Monitoring approach being rolled out across the UK as the successor to the EU Common Agricultural Policy. This means they will be able to monitor compliance and effectiveness of new policies right from the outset. The data products themselves represent value in the form of the customised, rigorously tested algorithms developed and implemented, along with a robust quality control regime, providing data that can be relied upon and can be verified with physical observations in the field. The markers and scheme assessment outputs scale reliably to address the Impact Potential, Technical Feasibility and Financial Viability of the monitoring.

FANTOM can promote the integration of ecological knowledge about the landscape into a data product, and model the change of the Habitat Quality and Biodiversity Potential so that customers can consider how a scheme is influencing changes over time and integrate this into policy development.

Current Status

The FANTOM project kicked off in March 2025. The first phase of the project is focused on consolidating the requirements baseline through workshops with key end users and stakeholders. The workshops explore the technical, performance, timeliness, quality and product format requirements.

HABTRAIL

Objectives of the Product

HABTRAIL addresses the challenges posed by Harmful Algal Blooms (HABs), which impact aquaculture, public health, tourism, and insurance sectors. HABs contaminate seafood, disrupt aquaculture operations, cause economic losses, and pose risks to human health. Traditional monitoring methods rely on costly, time-consuming water sampling and lab tests, often leading to delayed responses.

HABTRAIL provides an AI-driven early warning system that accurately predicts HAB occurrences and bivalve toxicity up to three days in advance. The platform integrates satellite and in situ data with advanced machine-learning models to deliver real-time insights.

Users access the predictions through a web dashboard, where they can visualise HAB severity, receive species-specific interdiction forecasts, and configure personalised alerts. Aquaculture operators can optimise harvesting schedules, public health offices can make informed safety decisions, and tourism and insurance entities can manage risks proactively.

By offering predictive analytics and automated alerts, HABTRAIL shifts HAB management from reactive to proactive, reducing financial losses, improving food safety, and ensuring sustainable marine resource use. The platform’s expansion will enhance its accuracy, cover more geographic regions, and incorporate additional environmental factors to support decision-making and regulatory compliance further.

Customers and their Needs

HABTRAIL targets aquaculture entities, national regulatory bodies, public health offices, tourism businesses, and insurance companies. These customers face significant challenges due to Harmful Algal Blooms (HABs), which disrupt operations, endanger public health, and cause economic losses.

  • Aquaculture entities (fish and bivalve farms) rely on HABTRAIL’s predictive insights to minimise losses by adjusting harvesting schedules and ensuring seafood safety. Their challenge lies in obtaining timely, cost-effective, and reliable data to mitigate production risks.
  • National regulatory bodies oversee seafood safety and environmental monitoring. They need accurate, real-time data to issue interdictions and ensure compliance, but current monitoring methods are reactive, costly, and slow.
  • Public health offices must prevent foodborne illnesses from toxic seafood. They require predictive tools to safeguard consumers and optimise response measures.
  • Tourism businesses (beach operators, tour agencies) suffer from reduced visitor numbers when HABs affect water quality. They need timely forecasts to adjust plans and avoid reputational damage.
  • Insurance companies rely on environmental risk assessments to price policies and process claims. A lack of reliable historical data makes risk evaluation difficult.

HABTRAIL directly involves these stakeholders by providing an AI-powered web dashboard with real-time predictions, automated alerts, and risk assessments, enabling data-driven decisions to mitigate the impact of HABs.

Targeted customer/users countries

HABTRAIL initially targets Portugal for deployment, focusing on its coastline and estuarine zones. Expansion is planned for Spain, France, and Italy in Europe (Year 2), followed by the USA and Chile in the Americas (Year 3), and later Japan and South Korea in Asia (Year 4-5). These countries have significant aquaculture industries, regulatory bodies for marine monitoring, and coastal tourism sectors affected by Harmful Algal Blooms (HABs).

Product description

HABTRAIL is an AI-powered web-based platform that predicts, monitors, and issues early warnings for Harmful Algal Blooms (HABs). It integrates satellite data with machine learning models to forecast HAB severity and bivalve toxicity with over 90% accuracy for the next three days. Users interact with the platform through a web dashboard, where they can visualise HAB severity maps, access species-specific interdiction forecasts, and configure custom alerts via email or SMS.

Innovative Aspects
  • AI-Driven Predictions: Unlike traditional reactive monitoring methods, HABTRAIL uses deep learning to analyse satellite imagery and identify patterns in HAB development.
  • Multi-Algorithmic Approach: Combines clustering, pixel segmentation, and deep neural networks to improve detection accuracy and reduce false positives.
  • User-Friendly Interface: Provides real-time, interactive dashboards tailored for aquaculture operators, regulatory bodies, public health officials, and insurers.
  • Scalability & Regional Adaptation: The model is designed to expand to new coastal areas with minimal additional data collection.
User Interaction
  • Aquaculture entities optimise harvesting schedules and reduce economic losses.
  • Regulators receive interdiction forecasts and cross-validate decisions.
  • Tourism and insurance entities monitor risk levels to adjust policies and business operations.

By shifting from a reactive to a proactive approach, HABTRAIL ensures safer, more sustainable management of marine resources.

Added Value

HABTRAIL brings significant added value compared to traditional monitoring methods and existing competitors by offering a proactive, AI-driven approach to Harmful Algal Bloom (HAB) prediction.

Key Differentiators
  1. AI & Multi-Algorithmic Analysis – Unlike competitors that rely on threshold-based methods, HABTRAIL uses deep learning to detect patterns in satellite imagery, reducing false positives and increasing predictive accuracy.
  2. Real-Time & Predictive Alerts – While current monitoring systems rely on costly, delayed lab testing, HABTRAIL delivers three-day advance forecasts, allowing aquaculture businesses and regulators to act before contamination occurs.
  3. User-Centric Dashboard & Automation – Unlike government-provided raw data portals, HABTRAIL offers a visual, user-friendly dashboard with customisable alerts, making it accessible to non-experts.
  4. Regional Adaptability & Scalability – Many existing solutions are site-specific or limited to post-event analysis. HABTRAIL’s AI models are scalable and adaptable to new coastal areas without requiring extensive new data collection.
  5. Comprehensive Stakeholder Support – Unlike competitors that focus solely on aquaculture or research, HABTRAIL supports national regulators, public health, tourism, and insurance sectors, integrating multiple industry needs into one solution.

By transitioning from a reactive to proactive strategy, HABTRAIL minimises financial losses, improves food safety, and enhances environmental protection better than any existing alternative.

Current Status

HABTRAIL is in an advanced development phase, with key achievements including the successful implementation of AI-driven predictive models, achieving over 90% accuracy in forecasting HAB severity and bivalve toxicity. A functional web dashboard has been developed, providing real-time data visualisation and customisable alerts.

Current work focuses on expanding geographic coverage from the initial L7C2 (Algarve) test zone to the entire Portuguese coastline, refining species-specific interdiction forecasts, and enhancing model precision using additional environmental data.

The User and Technical requirements were validated by Aquaculture Entities and Tourism Entities, gathering feedback for system optimisation, and integrating new features such as pixel-based HAB, SST, Salinity and CHLOR_A predictions, as well as Pollution, Sediments, Tide Levels, and Beach Information. Future expansion plans target Spain, France, and Italy in the next phase.

The project is progressing on schedule, with ongoing validation efforts moving on to the System Design and Data Integration phase, ensuring operational readiness for commercial deployment.

HySpex MSC

Objectives of the Product

The Methane Satellite Camera (MSC) from HySpex addresses the urgent need for accurate and cost-effective monitoring of methane emissions. Methane is a highly potent greenhouse gas, and is the second greatest contributor to global warming, only second to carbon dioxide. To limit emissions of methane, upcoming regulations demand effective emission tracking. By optimising our camera for satellite use, we will provide a reliable solution for industries to identify and reduce methane leaks at large scales, contributing to global efforts to mitigate climate change.

HySpex MSC allows for rapid deployment of methane detection capabilities as it aims to be a space-proven commercial-of-the-shelf solution, building upon our decades-long experience building hyperspectral cameras.

Customers and their Needs

Target customers for HySpex MSC are Earth Observation (EO) satellite service providers within the energy and environmental sector that operate their own satellites and constellations.

Currently, if a satellite service provider wants to offer methane detection as part of their services, their only option is to design and build their own instrument capable of detecting methane. This is a complex task requiring substantial financial investments over several years, involving multiple stages of critical research, development, and testing. The process carries a high level of risk, with no guarantees that the system will work. Such development typically falls outside the core competencies of these companies, and they would prefer to avoid such high-risk projects. Our HySpex MSC provides a cost-effective, ready-to-use solution that eliminates the need for service providers to develop such custom instruments, thereby reducing both financial risk and time to market.

Targeted customer/users countries

Global, but with specific focus on Europe and the US, who implement strict methane regulations.

Product description

HySpex MSC is a near-commercial off-the-shelf (COTS) instrument designed for methane detection from satellite platforms. Within the broader system and service ecosystem of its target users, HySpex MSC enables customers to launch and operate satellites capable of delivering high-resolution methane emission maps over large areas. This empowers end-users with actionable data for monitoring and mitigating methane emissions.

HySpex MSC incorporates our proven proprietary designs to achieve its mission. At its core it is a cutting-edge spectrograph design, integrated with a newly developed telescope and a diffraction grating optimised specifically for methane detection. Together, these components form a highly stable push-broom camera architecture.

Key performance highlights include:

  • Ground Sampling Distance (GSD): 60 meters
  • Spectral Sampling: < 2.5 nm in the spectral range of 1340–2500 nm
  • Signal-to-Noise Ratio (SNR): Optimised for precise methane detection

This innovative combination delivers unparalleled accuracy and reliability in methane mapping, while providing a scalable, efficient solution for satellite missions.

Preliminary overview of the HySpex MSC
Added Value

HySpex MSC is the first commercially available, off-the-shelf (COTS) hyperspectral camera capable of detecting methane from space. This innovative instrument empowers less integrated Earth Observation (EO) service providers to access methane monitoring capabilities without the need for complex in-house development.

Currently, providers without such access face an overwhelming challenge: building their own methane detection payload. This process demands substantial financial resources, years of research and development, rigorous testing, and carries significant risks, with no assurance of success. HySpex MSC eliminates these barriers by offering a ready-to-deploy, low-risk alternative.

By bridging this gap in the EO value chain, HySpex MSC accelerates the adoption of methane monitoring globally. This directly supports efforts to combat climate change by enabling broader and faster deployment of EO services for methane tracking.

HySpex MSC’s streamlined approach significantly reduces costs, risks, and time-to-market, delivering unparalleled value over in-house development while fostering innovation across the EO industry.

Current Status

Building upon the existing SWIR640 camera, HySpex MSC leverages HySpex proven spectrograph design, allowing for a short development cycle of only 12 months. Optical design and mechanical design of the new camera is well underway and will be completed early in the first quarter of 2025.

terrAIntel

Objectives of the Product

terrAIntel addresses the key challenges faced by non-technical users, particularly in the field of journalism. These challenges include technical barriers to accessing EO data products, limited interpretability without specialised expertise, processing delays that hinder timely reporting, and a lack of transparency in data sources and methodologies.

To overcome these obstacles, terrAIntel integrates Natural Language Processing (NLP) capabilities, allowing users to pose geospatial questions in natural language and receive intuitive, data-driven responses. This innovation has developed a demonstrator that transforms dense, complex EO data products into actionable insights, empowering non-expert users to confidently utilise EO data products for fact-based reporting, environmental monitoring, and urban development while minimising on-ground risks.

Key features include a user-friendly dashboard for data visualisation, customisable fields of interest, and metadata transparency to foster trust in the insights provided. terrAIntel streamlines the entire EO data product lifecycle by accessing diverse datasets and delivering tailored, comprehensible outputs for demonstration. This approach not only simplifies data interpretation but also promotes sustainability through improved data reusability.

Customers and their Needs

terrAIntel primarily serves non-experts, especially journalists, seeking intuitive access to EO data products. Key users include digital and investigative journalists, data storytellers, and, in the future, corporate users in fields like urban planning, disaster management, and environmental monitoring. These users often lack the technical skills to navigate complex EO data products or interpret outputs effectively.

By enabling users to ask geospatial questions in natural language and providing actionable insights through visualisations like maps and graphs with transparent metadata, terrAIntel addresses challenges of accessibility, interpretability, and timeliness. This allows users to make informed decisions without requiring specialised training. Feedback and pilot testing from journalists and media organisations ensure the platform meets their needs for accurate reporting and storytelling.

Targeted customer/users countries

The targeted customers and users of terrAIntel are global, with early adopters based in Austria, South Africa, and North America. These hubs serve as hubs for extensive networks of data journalists focused on digital media. The platform’s scalability and broad applicability ensure accessibility to global markets, supporting diverse industries while advancing the vision of expanding EO data usage among everyday users.

Product description

terrAIntel provides an intuitive, user-friendly interface powered by Natural Language Processing (NLP), allowing users to query EO data in natural language. What sets terrAIntel apart is its ability to grant access to EO data products and transform complex datasets into actionable intelligence through automated interpretation, complemented by visualisations such as maps and graphs.

This approach reduces the need for prior training and makes data interpretation more accessible, significantly enhancing the usability of EO data across industries. Key features include customisable fields of interest, real-time data access, and transparent metadata, fostering trust and ensuring accuracy.

Preliminary HAPS stratospheric segment design
Added Value

The EO industry faces a significant challenge with an overwhelming influx of data that is difficult to manage, access, and reuse effectively. This often results in EO data products being created as one-off outputs and subsequently forgotten. terrAIntel addresses this problem by enabling a broader range of users to interact with EO data without requiring extensive prior training. By improving accessibility and usability, terrAIntel empowers non-experts to derive meaningful insights, expanding the industry’s user base and promoting long-term data sustainability.

What distinguishes terrAIntel from competitors is its dual focus on both accessibility and interpretability of EO data products. While many solutions emphasise data access, terrAIntel goes further by helping users fully understand and utilise EO data. Through Natural Language Processing (NLP), automated interpretation, intuitive visualisations, and interactive queries, terrAIntel transforms complex datasets into actionable outputs, reducing reliance on specialised expertise and making EO data approachable to a wider audience.

Current Status

The terrAIntel project officially started with the Kick-Off Meeting on 9 December 2024. Upcoming priorities include gathering user requirements and refining the technical objectives in preparation for the Requirements Review milestone.

Skyfora’s Tropospheric Satellite

Objectives of the Product

National weather services and private weather companies can get many-fold more data than today from each weather balloon launch. These data contribute to more accurate weather forecasts.

This Tropospheric Satellite measures wind (including gusts and turbulence), standard pressure, temperature, and humidity variables. The main added feature is Airborne Radio Occultation, giving on average about 10 extra sensor-grade vertical profiles of temperature, humidity and pressure. As extra variables, it also measures infrared, visible light and UV-B radiation, and, optionally, air quality. The tropospheric satellite is standardly sold as a hardware product.

We also sell higher-level processed meteorological data. As a particular feature, AI and data fusion algorithms are being developed during the project. Data are processed with machine learning-based data fusion from other weather data sources, notably EO satellite data, to give more accurate weather data than what it is accomplished with Airborne Radio Occultation alone.

Customers and their Needs

The solution serves both public and private sectors, targeting national weather services (including storm measurement operators and climate/weather researchers), as well private weather instrument companies. The fundamental value proposition is simple: obtaining more upper-air weather data at very little extra cost.
In other words, meeting the urgent demand for filling important upper-air data gaps by using infrastructure and weather balloons that would be launched anyway. This leads to better weather forecasts and multiple societal benefits.

Targeted customer/users countries

The solution will be piloted with customers from Europe and USA.

Product description

The main building blocks are:

  • the StreamSonde radiosonde/dropsonde
  • an embedded GNSS receiver
  • a Ground Segment for telemetry and data processing
  • a Tropospheric Satellite Data Processing module, whose most important part is the radio occultation engine
  • an AI engine doing data fusion of tropospheric satellite and space-based satellite EO data.

The role of the Tropospheric Satellite, in the context of the overall system of its target users, is as part of a weather observation network, typically comprising a diverse suite of different instruments. The main new feature developed for the hardware is the ARO capability from the existing GNSS receiver.

The dual-band GNSS antenna (dipole) used in the StreamSonde/Tropospheric Satellite is excellent for ARO from weather balloons, as the main lobe of the antenna has a maximum in the horizontal directions, capturing low-elevation signals with maximum signal-to-noise ratio. Typical tilt angles for the Tropospheric Satellite attached to the balloon with a string are +/- 30 degrees during ascent. These angles are still within the main lobe of the antenna (antenna gain 0 to 1.72 dBi).

Added Value

With our solution, it is possible to obtain more upper-air weather data at very little extra cost. RO profiles from LEO satellites are worth at least €5 per profile. The ‘remotely sensed profiles’ are worth clearly less per profile than actual radiosonde profiles, but when adding all the extra profiles up, we get at a minimum of about €50 in commercial benefits for essentially the same radiosonde unit costs after an initial R&D investment. Typical studies also show that €1 invested into weather technology is returned to €5-7 of societal benefits, making the overall societal value greater than the commercial sales value.

By optimising the balloon, the Tropospheric Satellite can float for longer periods of time, even up to several days, further multiplying the benefits. Longer-living stratospheric or high tropospheric balloons may multiply the factor.

Moreover, as the Tropospheric Satellite’s data is processed with state-of-the-art GNSS processing algorithms and AI and data fusion methods, the accuracy and fraction of high-quality data will improve. Using advanced pattern recognition algorithms, refractivity profiles can be intelligently decomposed to their temperature, pressure and water vapor constituents, in a more precise way when using external EO data, and deeper into the troposphere than is currently possible.

Current Status

The Tropospheric Satellite project has successfully completed the Requirement Review, marking a key milestone in aligning mission objectives, system specifications, and user needs. It is now advancing into the development work packages within the de-risking phase. This stage will focus on the technical implementation and validation of critical components to ensure performance and feasibility ahead of full-scale development.

VALUESAFE

Objectives of the Product

VALUESAFE fills an important gap for companies and public institutions responsible for buildings and properties by helping communities and businesses understand and prepare for the financial impact of natural disasters like earthquakes, landslides, and floods.

VALUESAFE faces the challenges of risk and losses prediction by using satellite data, maps, and digital images. It offers a faster, accurate, affordable, and cutting-edge assessment of building vulnerability leveraging satellite data and AI, reducing reliance on slow, costly inspections. Through a simple online platform, users can request custom risk analyses and receive detailed reports that estimate possible damage and loss in value, making complex data accessible to a wider audience. These data help them make informed decisions about insurance, maintenance, and disaster preparedness, ultimately protecting their assets. By combining many data sources, VALUESAFE aims to make advanced risk assessment understandable and actionable, leading to safer, more resilient properties and communities.

Customers and their Needs

VALUESAFE is designed for public and private groups responsible for managing and protecting different building stocks, including cultural heritage sites. Private clients, like insurance companies and asset managers, need reliable, science-based risk data to guide business strategies and reduce unexpected losses. Meanwhile, public agencies can use VALUESAFE’s insights to prioritise mitigation actions, allocate resources, and plan safer communities. Both groups struggle with finding affordable, accurate risk evaluations that don’t require time-consuming, costly site visits. VALUESAFE’s online platform addresses these needs by offering tailored, easy-to-understand reports and a straightforward interface. This setup allows users to access vital risk information and connect directly with experts, helping them focus on long-term planning and resilience while saving time and resources.

Targeted customer/users countries

Italy and Europe.

Product description

VALUESAFE provides an online portal that enables users to access services, customise risk assessment requests, and manage outputs related to property vulnerabilities, risk and expected depreciation of buildings against natural disasters (earthquakes, floods, and landslides).

VALUESAFE is established on three tiers (Explore, Insight, Detail) based on increasing levels of accuracy and precision to target different users’ segments. To achieve this, VALUESAFE integrates three core evaluations: vulnerability assessment, analysing building characteristics that influence risk levels by employing digital and satellite imagery; hazard assessment, utilising recognised maps and satellite data to determine the probability and intensity of events; exposure evaluation, which measures the number of people, economic value, and cultural significance of assets in affected areas using national census data and expert insights.

VALUESAFE synthesises this information into accurate risk estimates and economic impact scenarios. This scalable solution provides crucial insights for diverse users and territorial extents – from public administrations managing large regions or critical infrastructure (schools, hospitals, heritage sites) to private insurers and investors – enabling informed decisions on asset management and disaster preparedness.

Added Value

VALUESAFE delivers added value through its reliable, expert-certified analyses that standardise vulnerability assessments on a global scale. By integrating diverse data sources, VALUESAFE provides a comprehensive understanding of building vulnerability, saving time and resources while delivering robust, replicable analyses directly applicable to business and regulatory decisions.

Unlike competitors that rely on outdated or non-site-specific data, VALUESAFE utilises advanced satellite data, interferometric analysis, and crowdsourced ground monitoring to produce up-to-date evaluations of multi-risk vulnerabilities across various real estate assets. This unique approach ensures that stakeholders receive the most relevant and timely information for informed decision-making.

Additionally, VALUESAFE’s user-friendly digital platform enhances this value proposition by allowing users to customise analysis targets, detail levels, and outputs to meet their specific needs. This flexibility and responsiveness ensure timely, detailed insights into risk and economic impacts, particularly for municipalities with rich cultural heritage, supporting vital economic development.

By focusing on a comprehensive approach, advanced technology integration, and user-centric design, VALUESAFE distinguishes itself in the competitive market for risk assessment solutions. This positions the service as a leader, driving significant progress in climate resilience and sustainable development efforts while enriching the overall customer experience and fostering long-term partnerships.

Current Status

VALUESAFE successfully passed the Requirement Review milestone on November 2024, and the proposing team is currently developing the design and development activities. Based on the feedback collected following interactions with stakeholders, the architecture of the system has been defined to effectively integrate multi-source information in comprehensive risk assessments and financial impact evaluations. The effectiveness of the proposed solutions will be soon tested in real case study applications focusing on representative geographical clusters and built environments.   

SPIRIT

Objectives of the Product

A significant gap exists in current EO data availability. Most infrared imagers in space offer images with a resolution of 60 to 100 meters per pixel, capturing images of specific areas only every two weeks. Our objective is to address this gap by providing high-resolution (~10 meters per pixel) thermal imaging with rapid global coverage. This combination has enormous potential to monitor progress and inform decision-making to combat climate change. By directly measuring heat emissions in the built environment, our frequent high-resolution thermal imaging precisely identifies where and when significant energy inefficiencies occur. This empowers governments, businesses, and individuals to take targeted actions to improve energy efficiency, crucial for transitioning to a net-zero society and meeting sustainability targets.

Additionally, our thermal imaging helps identify heat-stressed zones and indoor overheating during heatwaves, reducing cooling energy demands. Previous services like this were costly due to the need for large telescopes, but recent advancements in telescope technology and the rise of low-cost nanosatellites make our solution affordable and commercially viable as we transition to a net-zero economy.

Customers and their Needs

Our space telescope will help EO satellite operators who want to capture high- resolution thermal infrared satellite images by reducing development costs and enabling large satellite network.

Targeted customer/users countries

Main Locations: UK, Europe and US
Interested and prospective customer companies include: OpenCosmos, Jet Propulsion Laboratory (JPL), Carbon Laces.
Prospective Customers in the Space sector: Planet (USA), Maxar (USA), Airbus (EU), Blacksky (USA), Satellite Vu (UK), ConstellR (Germany).

Product description

We are developing a powerful thermal infra-red (TIR) telescope. It has very high resolution (6.5 metres per pixel), a very large swath (33 km) and it can collect data in a push-broom mode at a very high rate (940,000 square kilometers per hour). Its low unit cost means that large constellations are feasible, which offer daily revisit rates. The customer will buy the product and integrate it into a satellite platform. The satellite will capture EO data which will be sold to the end users.

The product includes an Image Following System (IFS), which removes the blurring caused by the ground motion (~7 km/sec).

Added Value

Our resolution in the TIR band is much better than current or planned offerings. For example, the resolution of LandSat is about 60 m per pixel. SatVu (HotSat) offer a similar resolution to ours, in the MWIR, but they do not offer a large swath, so their data collection rate per satellite is lower..

Current Status

SuperSharp has been developing TIR space telescopes for several years. Part of this development is the use of uncooled micro-bolometer arrays (UMBAs), which are relatively unused for space applications. SuperSharp now has expertise in this area.

The optical design of SPIRIT has been completed. It is innovative because it has a very large field of view (5 degrees), a large aperture (55 cm) and is very compact and lightweight. We have also already started developing an image following system, which enables strip-mapping (push-broom) data collection.

SmartDig

Objectives of the Product

SmartDig, a service based on AI, provides accurate preventive archaeology, a legal requirement for businesses and organisations undertaking construction affecting ground. Detection of surface and sub-surface features helps Cultural Heritage professionals extracting information about potentially buried archaeological remains, reducing waste of time, money due to unplanned delays and avoiding the loss of
valuable historical structures.

Customers and their Needs

The targeted customers for the SmartDig service include professionals involved in public or private construction projects, such as project managers, architects, engineers, and urban planners, who must comply with legal obligations to safeguard Cultural Heritage (CH) and archaeological sites. These stakeholders are involved in various stages of project development, including the design, planning, and execution phases. Their primary need is to ensure that any archaeological features in the area are preserved without causing significant delays or cost overruns to their projects.

Current challenges include the time-consuming process of non-invasive archaeological analysis, which can take up to six months and increase project costs. This delay is due to the extensive retrieval and analysis of photographic and satellite data. Moreover, the complexity of navigating legal frameworks, such as the “Verifica Preventiva dell’Interesse Archeologico” (VPIA), adds further constraints.

SmartDig addresses these challenges by offering a faster, more efficient service for conducting preventive archaeology analyses. By simplifying access to crucial geospatial and archaeological data, SmartDig helps professionals reduce the time and cost associated with the preliminary assessments, enabling smoother project execution while maintaining the integrity of CH sites.

Targeted customer/users countries

The SmartDig service initially targets customers and users in Italy, where legal requirements like “Verifica Preventiva dell’Interesse Archeologico (VPIA)” ensure the protection of Cultural Heritage (CH) during construction projects. However, the service is designed to be adaptable and can be applied across Europe and the rest of the world. It is relevant for any region committed to preserving CH and archaeological features during development activities. Many countries have similar legal frameworks or guidelines for protecting heritage sites, making SmartDig valuable for professionals in public and private sectors globally. The service can assist any country or region seeking to streamline the archaeological assessment process while maintaining a strong commitment to CH preservation.

Product description

SmartDig is an innovative, GIS-based web application designed to streamline preventive archaeology. The tool leverages Artificial Intelligence (AI) and Earth Observation (EO) data to conduct multi-temporal analyses, identifying potential buried archaeological features. This enables SmartDig to detect vegetation, soil moisture, and surface anomalies—such as crop marks, soil moisture marks, and micro-relief—that indicate buried remains.

One of SmartDig’s key innovations is its multi-temporal analysis, which enables users to track seasonal changes and compare data over time, improving the accuracy and reliability of anomaly detection. This system replaces traditional, manual methods, providing results more quickly and consistently while reducing errors. The service offers spot-on-demand analysis or subscription-based access to recurrent analyses, as well as API integration for expert users needing seamless integration with GIS or EO tools.
Users interact with SmartDig through an intuitive web platform. They can upload their Area of Interest (AOI), customise the analysis (e.g., spatial resolution, accuracy), and receive results in an accessible format. SmartDig reduces the time needed for preventive archaeological analysis, offering a more accurate, non-invasive, and cost-effective solution to safeguarding cultural heritage during construction projects.

Added Value

SmartDig brings substantial added value compared to existing competitors by addressing the limitations of traditional preventive archaeology methods. Competitors often rely heavily on manual inspections or high-resolution drone-based imagery, which, while detailed, are costly, time-consuming, and limited in coverage. Drones may provide excellent resolution in small areas but struggle with large-scale projects due to the time and expense required to survey vast regions.
SmartDig, on the other hand, uses AI-driven analysis of Earth Observation (EO) data from satellites, enabling rapid, large-scale, multi-temporal analysis. This approach not only allows for the identification of a wider range of archaeological features, such as crop marks and soil moisture marks, but also minimises the need for costly and slow on-site inspections. Additionally, SmartDig’s ability to analyse historical EO data offers a significant advantage over competitors that rely on real-time data alone, as it can detect changes over time that may indicate hidden features.

Current Status

The SmartDig project is progressing well with several key steps already achieved. We have successfully engaged multiple stakeholders, including archaeologists, researchers, and professionals in the construction industry, to gather in-depth user requirements. This engagement has provided valuable insights into user needs, allowing us to refine the service accordingly.

The preliminary analysis of the best AI architectures is underway, focusing on identifying models that can efficiently process EO data and detect archaeological features with high accuracy. Additionally, preliminary test sites have been selected—some areas with well-known hidden structures and various types of anomalies. These sites will serve as a crucial benchmark for testing the service’s performance.
We have also started assessing the best combinations of EO data for anomaly detection, evaluating different satellite data sources (e.g., LiDAR, SAR, optical) to determine the most effective options for identifying buried archaeological features. These efforts are laying the foundation for further development and testing.