The Earth Observation (EO) market is entering a transitional phase, in which diversified data streams are becoming available to the users from both governmental (traditional) and commercial actors. The oversupply of imagery constitutes an opportunity for demanding “data-hungry” applications, particularly the ones of the security and intelligence domain, as understanding the patterns of life of any observed area or event is a prerequisite for effective analysis. On the other hand, this opportunity poses various questions with regard to the current imagery and geospatial intelligence (IMINT/GEOINT) production capacities: How will the IMINT/GEOINT agencies deal with all these new sources? How can IMINT be scaled to catch up with the flow of imaging streams? What will be the impact on the operational production models? The paper discusses practical issues and challenges in pivoting the market towards IMINT analytics.
The operational IMINT production processes are typically task-based and organised as sequential series of manual “pipelines”. While this approach optimises control over the production chain, it is evident that the vast amount of available imagery will challenge the current production modality and will create the need for more automation and augmentation. The use of the so-called “platform” model for GEOINT, as proposed in Holmes et al. 2018, is briefly discussed with references to the recent developments in crowdsourcing, open-source data exploitation and cloud-native geospatial ecosystems. The shift towards IMINT analytics for security applications is radically connected with the developments in Artificial Intelligence (AI) and data fusion. I claim that the majority of the IMINT-focused AI solutions available in the market are not yet mature enough, as many of the problems to be tackled in the security domain are either task- or context-specific, requiring inclusive and careful training of algorithms. It is argued that the optimum AI approach for IMINT should involve a “symbiotic environment”, where AI tools are suggesting tags for features and activities; probably coupled with Natural Language Generation (NLG) systems. In this environment, analysts will act as validators/curators, continuously improving the algorithms through active learning.
Apart from the technological aspects, community-building activities are required to structure the demand of the security sector and deliver relevant and trustable offers. This is particularly important for vertical markets that are not yet fully exploited (e.g. law enforcement) and for which the project-based approaches seem to be delivering weak engagement results. Analytics can be the supportive element for risk analysis and intelligence investigations. The understanding of the operational environment and the capacity to deal with demanding requirements will be determining factors for successful market players.

On 16 September 2018, the NovaSAR satellite was launched from the Satish Dhawan Space Centre in India. NovaSAR is an experimental low-cost satellite mission developed and operated by Surrey Satellite Technology Ltd in the UK. It carries an S-band Synthetic Aperture Radar (SAR) and an Automatic Identification System (AIS) for ship identification.

Australia’s Commonwealth Scientific and Industrial Research Organisation (CSIRO) is a partner in the NovaSAR programme – together with the UK Space Agency (UKSA) and Indian Space Research Organisation (ISRO). CSIRO has purchased 10% of the satellite operational capacity, providing Australia with exclusive rights for acquisition programming and data downlink, as well as processing and distribution over its national territory and other designated areas of interest. For Australia, the NovaSAR partnership presents an opportunity to build new space R&D facilities and to stimulate new applications research in Earth observation, as well as to grow partnerships and potential commercial markets around this new capability. In the longer term it constitutes a pathfinder towards a potential full-scale national satellite constellation.

The NovaSAR radar operates in S-band (3.2 GHz/9.4 cm) which is a rather uncommon frequency in between L-band (23.5 cm) and C-band (5.6 cm), and therefore of particular scientific interest for both terrestrial and maritime applications as a complement to other operational SAR missions. Through its partnership in the NovaSAR mission, CSIRO will be demonstrating that even with an experimental mission with limited acquisition capacity – NovaSAR has a payload duty cycle of 2 minutes – it is possible to optimise acquisition planning and both build a multi-seasonal continental-scale archive of time-series data, as well as to support an international research programme with high-temporal observations over dedicated study sites. Initial focus is all Australian (EEZ) landmass and Oceania, and subsequently in potential collaboration with other mission partners and the international science community, regions and study sites in other areas of the world.

NovaSAR data from CSIRO will be publicly open and distributed free of charge.

The aim of this paper is to show how synthetic aperture radar (SAR) data can be used to support operational ship routing. We describe end user needs and what information that can be retrieved from SAR. An evaluation of the possibility to provide information derived from SAR for long distance ship routes is presented. The EONav ship routing service is used to demonstrate the operational feasibility.

In short, for safe sea transport, the ship master wants to have information on unfavourable weather and ice conditions that should be avoided. This information has traditionally been provided by weather forecasts and ice charts. With information on currents, wind and waves it is also possible to select routes that are faster or more fuel efficient. Route selection has to a large extent been based on previous experience of the conditions in the geographic area. SAR can provide information about wind, waves and sea ice conditions, while the possibility to estimate ocean surface current is limited to methods that are still severely restricted by low availability of suitable SAR data.

The availability of SAR data has increased rapidly over the last years, especially with the launch of the Sentinel-1A and 1B satellites and the Copernicus open-data access policy. This has increased the possibility to utilize SAR data for operational services to the maritime sector. With access to a number of planned ship routes we have performed an evaluation of the possibility to acquire near real time SAR data for these routes. This evaluation clearly shows that Sentinel-1 data must be complemented by SAR data from other missions. The example in the attached figure shows that for a two week route across the Pacific Ocean in December 2017, the Sentinel-1 only provided one image that covered the route, while large parts of the route could be covered by near real time data ordered from Radarsat-2 and COSMO-SkyMed. To assess the usefulness of the SAR data, wind speed retrieved with two different SAR algorithms have been compared with re-analysis data from the ERA-5 model, wind products based on the ASCAT scatterometer data and wind observed by ship sensors. The results indicate significant differences between the different sources of wind information, which emphasize the importance of access to information from several independent sources.

The EONav ship routing service uses information about ocean surface currents, waves, wind, bathymetry and sea ice conditions to plan the optimal route for a vessel. Products from the Copernicus Marine Environment Monitoring Service (CMEMS) and several national weather services are used together with in-house products developed for SAR data from satellites such as Sentinel-1, Radarsat-2 and COSMO-SkyMed. A sail plan optimized e.g. to reduce fuel consumption is determined based on the available met-ocean information and is communicated to the ships to assist captains in selecting the best route and speed pattern to their destination. Metocean information based on SAR data can be used for training and validation of the system, but methods to incorporate near real time SAR products are also considered.

Figure caption: SAR coverage for one of the ship routes across the Pacific Ocean. The figure shows the period from the 15th to the 31st of December 2017. Green frame = Radarsat-2, red frame = COSMO-SkyMed, blue frame = Sentinel-1.

The extraction of oil from the Alberta Oil Sands often relies on thermal recovery methods, which warm and reduce the viscosity of the bitumen. The most common thermal recovery types are Steam-Assisted Gravity Drainage (SAGD) and Cyclical Steam Stimulation (CSS), both of which use steam to transfer heat to the reservoir. A common side effect of these steaming operations is the expansion of the reservoir and the uplift of the ground surface above the steamed area. Alberta is at the forefront in the environmental monitoring of thermal recovery operations and promotes the use of advanced remote sensing approaches such as InSAR to monitor ground deformation.
Thermal recovery operations in Alberta have now been in use for many years and some sites are starting to approach their end-of-life. The regulator in Alberta has expressed interest in monitoring the complete life cycle of a thermal recovery operation to better characterize their environmental impact, including any residual surface deformation. InSAR is in a unique position to provide the complete deformation history of SAGD or CSS sites as archives of SAR imagery date back to the early 1990s.
The Canadian Space Agency (CSA) has funded a project to not only process these archives but to also promote research aimed at creating complete time histories of deformation using satellite archives acquired from different sensors, at different times, from different orbits, with different viewing angles and different wavelengths. The final objective is twofold: completely characterize the ground deformation history of a thermal recovery project from start to finish while also advancing the state-of-the-art in terms of InSAR processing.
This latter objective focuses on the fusion of multiple past SAR archives with data from current and upcoming SAR missions. Today normal InSAR practices involve the separate processing of individual SAR data sets to derive measurements of ground deformation. As satellites have a finite lifespan and subsequent SAR missions often have different characteristics, it is not possible to perform a coherent processing of the individual SAR archives to produce uninterrupted time series of motion spanning the entire period covered by the multiple satellite missions. We propose to overcome this limitation by developing an advanced InSAR fusion technique to exploit the lengthy historical archives of all past and present C- and X-band satellites and create continuous, uninterrupted vertical deformation time-series.
The presentation will explain the methods used and how InSAR data of this kind can support regulators with case studies and examples.

In the commodity markets, accurate and timely information is required for effective decision-making by analysts and traders. Data on mining activity and production is widely available but not always accurate or up to date, and often self-reported by producers. Commodity traders want to receive timely and regular insights on production levels to guide their trading activities, particularly for regions of the world where information is not readily available. They want to improve their understanding of the correlation between market price movements and the physical production and movements of commodities in the world’s supply chains. Traders want information that is easy to access and use, and want to understand how that information was created and what it means.

Earth-i is developing the SAVANT service, in the frame of an activity co-funded by the European Space Agency through its Business Applications Programme, which is an end-to-end analytics and insights service that provides copper traders with a new information source drawing upon data gathered by Earth Observation satellites. There is a subset of mines around the world that significantly influence the supply and price of copper; monitoring these regularly with high-resolution satellites provides a valuable stream of observation data. SAVANT then uses advanced analytics to combine this up-to-date satellite imagery with complementary terrestrial data sources, to generate and deliver products tailored to the specific information needs of the commodity traders.

In particular, Earth-i is using innovative volumetric and activity assessment to turn satellite imagery of the mines into deeper insights into copper production at these crucial locations. Correlating this with complementary data sources, such as published market and production data, and news feeds from mining company trading statements, SAVANT produces actionable data and insights for traders – available online and tailored to their specific needs.

In the future, Earth-i will also be able to extend the service using its Vivid-i commercial constellation of high-resolution, full-colour video satellites to provide high-revisit coverage of mining-related sites of interest. Video from space offers the unique capability to extract high-precision volumetric data from a single pass of a satellite over an area of interest, and provides the ability to observe and interpret activity at the mines by detecting moving features and objects, such as vehicles, trains and conveyor belts, to enhance the level of detail and insight provided by the service.

This presentation will describe the innovative aspects of the programme and discuss operational challenges and barriers Earth-i has encountered along the way, and the novel solutions Earth-i has put in place to overcome them. The results of initial commercial trials will be presented and the roadmap to commercial exploitation will be explained, including how the service will be extended to cover other types of mined commodities in future.

In Norway operational maritime surveillance based on satellite data started in 1991 with the launch of ERS-1. The services was over many years based primarily on SAR data from the European ERS-1/2 and ENVISAT missions and the Canadian RADARSAT-1 mission. Today the main SAR missions used in the services are Sentinel-1 and RADARSAT-2 with data from other commercial SAR missions as extra data sources when needed. These data sources are combined with AIS data from ground based receivers and the Norwegian AISSAT-1/2 and NORSAT-1/2. In the coming year data from the navigation radar detector which will be flying on NORSAT-3 and the new MicroSAR mission will be included in the services. The services in Norway is based on quasi-real-time concepts, to ensure that the end users get their information within very strict time requirements. Another important characteristic of the operations services is the sharing of data and results between different governmental agencies. This is done through the Barentswatch which is a portal for cooperation, professional development and sharing of information for governmental agencies and the public. Today the main satellite based services are related to ship detection, oil spill detection and sea ice monitoring.

The presentation will detail the functional and operational concepts and how the services are designed to fulfill requirements. A basic idea in the concept for the Norwegian services are the use of many data sources to provide a more complete situation awareness picture. This is done by fusing together information from all available data sources, to provide the situational information both in space and time. The services are today increasingly utilizing new earth observation and information technologies. Some example of usage of for instance machine learning will be presented in the talk. At the end of the talk a short summary of new operational capabilities, which will be included in the operational services the coming year will be given.