Print Friendly, PDF & Email

Geoinformation is the representation and analysis of data on a map by an information device. The increasing number of Earth observation satellites and the possibility of using artificial intelligence to manage and prioritise the information collected are driving their use for military and security purposes.

The demand of geoinformation related products is growing along with the development of geospatial intelligence applications (GEOINT). GEOINT is a predictive analysis method focused on monitoring given variables on a map over time. Information gathered through GEOINT are then presented to users on an interactive multi-layered map integrating a scatter-plots. The more detailed the scatter plots are, the more comprehensive is the situational awareness of the users with regard to the issue that they are investigating. GEOINT techniques have been used for decades, particularly by domestic security services in the US, in order to monitor the activities of large groups.

For instance, the National Security Agency’s (NSA) PRISM programme, which became famous with the ‘Snowden case’, is based on the gathering of metadata about words used. Given a normal distribution of words in an average conversation, the programme would automatically “red flag” conversations containing a higher number of occurrences in a given time frame. In case of a recurrence of these red flags over time, the NSA would initiate an investigation in the area where they occurred. Today, this principle can be applied to even more complex applications capable of monitoring multiple domains at once. Big data, sensors, satellites and other sources are so abundant now that multiple aspects of an issue can be monitored over time, thus providing the decision-maker with accurate situational awareness and even with some prediction capabilities based on behavioural analyses. The rise of artificial intelligence (AI) applications feeds the proliferation of geoinformation applications based on GEOINT methods. As these applications have large potential, the demand for geoinformation services is increasingly focusing on developing currently existing GEOINT tool sets while also developing more ambitious new ones as well.

Game Changers in GEOINT

The impact that data gathered through GEOINT apps have on decision-making is strongly dependent on the amount of information obtained and on their relevance over time. The increasing number of earth observation satellites and miniaturisation have been “game-changers” in the development of GEOINT, as they make the collection of spatio-temporal information more regular and more affordable. The 200 meteorological and earth observation satellites that have been launched in the last decade provide high-resolution, night and day monitoring of land, air and sea over a given time frame, providing continued observation of a given area. Once the user has designated the area of interest, observation satellites can be redirected to adapt GEOINT to the analysts’ needs. As the number of satellites is expected to increase consistently until 2024, reaching up to 1,700 pieces, 1700 pieces, the amount of information that will be collected by space sensors is expected to grow accordingly, thus enhancing GEOINT performances. In addition, the fact that a growing number of satellites have synthetic aperture radars (SAR) instead of classic electro-optic sensors served to strengthen image quality, as it allows for the collection of images irrespective of the weather condition – a crucial issue in the past, especially in geographical areas with harsh weather conditions. ASI’s COSMO-SkyMed constellation is one of the most remarkable examples of dual use space-earth observation systems. Consisting of four-medium sized satellites, each one featuring a microwave high-resolution SAR, the constellation has all weather/day and night acquisition capabilities. The large number of radar images that this constellation can gather are increasingly exploited for feeding GEOINT applications. This capacity is exploited both in-house, with products such as SEonSE (Smart Eyes on the SEa), and by external service providers (see also below).


Miniaturisation as well has been maximising the utility of information gathered through GEOINT applications, as it makes data flows more consistent due to applications in satellites and processors. The development and deployment of nano- and micro-satellites, over the last decade, is expected to rise in the coming years. As a result of miniaturisation, satellites are now much smaller than in the past and, therefore, cheaper in terms of construction and launch costs. Modularity in payloads given by products such as CUBESATs has made producing and launching satellites more affordable for a growing number of actors interested in GEOINT activities, from institutions to private companies. Even if some technical obstacles remain, such as propulsion and power storage, these satellites can be launched in large numbers to compensate for their limits. Moreover, despite the fact that image resolution is lower compared to bigger models, they could be used along with them to provide the best results possible. Indeed, nano-and micro-satellites can provide persistent even if not high-resolution images of an issue, which can then be better observed due to the information collected by more performing satellites in order to provide analysts with more comprehensive data.

Miniaturisation has also an important role when it comes to processors as it constantly increases technical capabilities in stocking and processing a vast amount of data. More performing stocking and calculating power raises GEOINT efficiency, as it allows for better populating maps and, consequently, to considerably raise users’ spatio-temporal situational awareness of the area of interest.

Adapting GEOINT to Metadata

The important contribution of satellites and miniaturisation to GEOINT in terms of data supply has multiplied the amount of information to be analysed. In particular, analysts have to cope with a large number of high-resolution images collected in near real time that have to be merged with relevant information collected through other tools (for example, social media) before they can be converted into meaningful data. However, dealing with such a flood of information is time-consuming. Although the digital tools that analysts rely on are becoming increasingly powerful, especially in terms of computing power, analysts remain at the core of raw data processing. The broader the spectrum of information to be combined, the more time and professionals are needed to deliver insights about the issue being analysed. The biggest drawback is that the advantage of capturing real-time information is lost because the delay between capturing information and providing analysis can be weeks or even months in more complex cases.

Digital technologies can help solve these problems by automating part of the analysis cycle and making it faster. This has led to the development of new Activity Based Intelligence (ABI) methods that include the use of Artificial Intelligence (AI).

ABI is a set of spatio-temporal analytic methods based on the integration of metadata collected through different intelligence gathering solutions. It helps analysts to overcome difficulties derived from the data’s variety, volume and veracity, allowing for analysing quicker elements “in motion”. ABI is an important support to GEOINT activities, as the rapid merging of information allows for discovering relevant patterns that could then be reshaped according to the changes identified to enhance the quality of intelligence gathering. The ABI deductive method allows analysts to maximise the widespread information gathering that modern technologies can bring, as they can adapt collection to their needs and rapidly merge consistent and heterogeneous data sets.

Applying AI to this process makes it even faster. For instance, AI software can rapidly, if not instantly, select useful information according to given criteria, or merge information on the same issue but gathered by different tools to show them on a unique thematic map. Moreover, AI can merge heterogeneous data sets into homogeneous outcome. The emergence of deep learning techniques relying on a neural network approach, coupled with hardware evolution, makes classification, detection and segmentation more accurate. By being applicable to several ABI needs, from very high-resolution satellites to constellations of micro-satellites, deep learning techniques make ABI more efficient.


AI is a push factor for the development of GEOINT-tailored technical services that transform raw information into usable data. Indeed, AI software can map and monitor several issues, from critical infrastructures to crisis management and maritime security. Not surprisingly, the demand for AI applications in the GEOINT domain is particularly strong in countries, who are heavily investing in AI Research and Development,namely Japan, China and US.

The US-based company Orbital Insight is one of the leading companies in geospatial analytics. As a result of its remarkable cloud computing, machine learning and AI tools, the company can provide cloud-based products that gather, process and transform multiple geospatial datasets. Orbital Insights, which is in charge for monitoring 85% of US refineries, offers several solutions for the public sector, for instance for the detection of new infrastructures and the characterisation of military activities.

Moreover, the US government is seeking to externalise data collection activities and libraries to better focus on analysis, thus encouraging the development of new solutions and software.

After having lagged behind, European countries have recently understood the importance of entering the “AI race”. Thus, dedicated investments in Research and Development have been increasing both at the national level (namely in France) and at the European level, in particular in the EU multiannual financial framework 2021-2027. AI applications to the intelligence cycle are part of this effort, with European companies already working on the development of dedicated software.

For example, the Italian company e-GEOS (a joint venture of ASI and Telespazio) with its service portfolio, which combines images from the Cosmo-sky med constellation with metadata from various unclassified data sets, is a leading provider in the field of geoinformation. The services will be offered in multiple versions to optimally adapt them to the application areas, including civil and military applications as well as applications for defence and intelligence services. To this end, e-GEOS offers the BraINT solution, a state-of-the-art modular IMINT analysis environment that integrates proprietary solutions with bespoke tools to support analysts throughout the intelligence cycle. The system can be used by defence ministries, defence agencies and security institutions for target analysis and/or monitoring, detection, damage assessment and mapping. Since 2018, the company has also provided a platform for the naval environment, Smart Eyes on the SEas (SEonSE), to provide users with meteorological and oceanographic information or the exact position of ships. Not surprisingly, E-Geos, in collaboration with Orbital Insight, is using cloud computing solutions and AI techniques developed by the US company to analyse high-resolution images collected by the Italian company.

On 24 April 2019, Airbus Defence and Space and Orbital Insight launched their partnership in the development of the Earth Monitor analytics software. This explains well how merging existing libraries and platforms with machine learning can maximise users’ knowledge, allowing them to allocate resources in a time-efficient manner. Earth Monitor is expected to track changes over pre-registered and new areas of interest through Airbus’ imagery archives and satellites capabilities and Orbital Insight’s dynamic algorithms that can detect changes in near real-time, for example, traffic analysis and detection of new infrastructures.

France will be one of the European countries who will invest the most in the development of AI technologies – €100M a year between 2019 and 2025, as the draft by Defence Minister Parly (dated 5 April) shows. In addition to large companies, such as Thales and Nexter Robotics, a dozen digital start-ups have specialised in AI applications in the defence sector. Two of them, Geo4i and EarthCube, are working on the development of AI tools for image analysis.

A European Approach

European collaboration in the space domain is one of the most remarkable success of the EU. Indeed, European countries have been able to build a system that could not only provide raw information but also perform the whole intelligence cycle due to the work of dedicated agencies.

For what concerns information gathering, European companies working in the GEOINT domain could rely on information gathered through one of the most ambitious Earth observation programmes in the world – Copernicus. Born out of the collaboration of the European Commission with the European Space Agency (ESA), this user-driven programme is aimed at developing European information services based on information gathered due to the Copernicus Sentinel satellites constellation (with the eventual support of Contributing missions) and in situ-sensors managed by the European Environment Agency.

Raw data collected by Kopernikus’ satellites will then be converted into usable information by six Kopernikus services with different environmental and security specialisations – land, sea, atmosphere, climate change, emergency management and security. As a result, the information collected under the programme helps decision-making in areas such as structural policies (for example, fisheries, agriculture and local planning), health, transport, civil protection and tourism.

To better serve EU foreign policy activities, the European Union Satellite Centre (SatCen) provides the European External Action Service (EEAS) with fast and reliable analysis of information gathered by earth observation, thus providing support to humanitarian and aid missions, contingency planning and general surveillance. SatCen also monitors and analyses critical infrastructures, military capabilities and weapons of mass destruction. Following the delegation agreement signed in 2016, the European Commission entrusted SatCen with the operations of the Copernicus Service in Support to External Action (SEA), the EU geo-intelligence service, which became operational in 2017. This provides relevant national and European stakeholders (EU Commission, EEAS and people serving the Common Security and Defence Policy) with increased situational awareness on crisis prevention and management.

Users who are interested in detecting and/or monitoring events outside the EU, but who have a potential impact on the EU, can request geo-information on demand, which will be analysed and delivered in tailor-made news products. Satellite images can be presented in various forms, from short reports to detailed thematic maps. The service can provide both continuous monitoring of the issue as well as rapid intelligence support in responding to crisis situations such as political or armed conflicts.

In April 2019, the European Commission launched a call for proposals for projects under the European Defence Industrial Development Programme, one of the instruments introduced last year to create a genuine European defence. As proof of the importance that GEOINT and related disciplines have for European countries, the ISR-PEO call focuses on improving persistent Earth observation from space through “automated interpretation of data and information, including artificial intelligence, cloud solutions and real-time on-board processing by sensors”.

The prototype to be developed by the contractor shall implement the ABI concepts by automatically extracting units from different data sources in order to gain insights from the aggregation, correlation and monitoring of such units with respect to patterns and anomalies. In line with these developments, the ABI systems should be applicable as a new tool for Intelligence, Surveillance and Reconnaissance (ISR) missions. The requirements also include support to GEOINT analysts in defining the most appropriate data planning and collection strategy to perform ABI tasks, in particular for sea and land applications. Due to its AI features, the ABI method is expected to reduce the workload of analysts for data preparation in order to focus on the contextual analysis of GEOINT.

Final Remarks

It is expected that GEOINT will gain in importance in the future. The multiplication of Earth observation satellites, including nano- and microsatellites, provides real-time information on the evolution of certain issue occurring in a given area over a period of time. This information can then be merged with that collected by other methods. Technological innovations help to cope with the enormous volume of information as they speed up the transformation into exploitable intelligence. If the ABI method allows better management of sources of complexity and uncertainties in data, the AI software can perform some tasks autonomously. These technological advantages have allowed analysts to focus on the most challenging tasks, bringing great benefits to the entire intelligence cycle.
This innovative approach to geo-information, which focusses on the analysis of data flows, improves the reaction time of decision-makers and thus realises the information superiority of users.

Giulia Tilenni is an analyst in international affairs based in Paris, France.