Building the Global Events Observer: Five questions with the European Space Agency

In August 2020, McKenzie Intelligence Services was awarded EUR 685,000 in co-funding by the European Space Agency (ESA) Space Solutions to build and deliver a Global Events Observer for the insurance industry, enabling the further collection and use of highly accurate, geotagged external data from a range of sources to provide early warnings of loss events.

We caught up with ESA Business Applications Engineer Volker Schumacher and Head of ESA Space Solutions Nick Appleyard together with MIS Founder and CEO Forbes McKenzie to discuss the project and progress so far.

How did the idea for the Global Events Observer (GEO) come about?

Forbes: The idea came from talking to our clients in the re/insurance sector about what they needed next from geospatial imagery, and the market told us they wanted better real-time intelligence and analysis on damage post catastrophic events. Using combined space and ground-based sensing capabilities the collaboration with ESA is intended to enable the global tracking of catastrophe timelines and delivery of user specific reports for Exposure, Claims management and Claims Reinsurance users in a scalable way. Essentially the market is looking for very early data from catastrophe or other loss events, delivered at far higher quality and quicker than it has been able to access before.

Nick: MIS approached us in 2018 with a pitch to enhance capabilities for identifying and tracking real-time damage to property and transportation infrastructure. We felt the idea for the GEO was a natural fit between the insurance industry, which uses data to quantify and manage risk, and the satellite EO sector, which generates detailed and wide-ranging data about what is happening anywhere in the world. So the GEO proposal brings those two things together to make excellent use of that fit.

Volker: We feel the GEO is a very good example of what satellite-based Earth observation is able to deliver on a 24/7 global basis, and it is constantly improving – there are increasingly more and more data providers coming online. One reliable Earth observation data source which is already fully operational and growing in size and diversity in offering an increasing range of different kinds of data is ESA’s Copernicus Sentinel fleet, for instance, making available data with a free and open data policy to businesses all over the world.

Can you tell us more about the technology behind the GEO (constellations etc)?

Forbes: There are two elements to the GEO, and both significantly enhance our global intelligence grid. The first is the use of historical data, the GEO houses data for every global peril since 1979, to allow pre and post-event comparisons. The second is a highly detailed real-time view of the current situation on the ground at any point on the planet using the various cutting edge geospatial and other data sources available. Both these elements are exciting of course but really come into their own when combined with the on the ground intelligence, machine learning and artificial intelligence programmes we use to analyse all this data and apply to our detailed understanding of how exposure management works.

Nick: The GEO service proposed by MIS draws on real-time on satellite-based Earth observation data as well as archived data for collecting pre- and post-event data. For this, MIS already uses multiple space based remotely sensed data sources and plans to expand as new sensors and capabilities are launched.

To this end, if you look into how the Earth observation industry has developed over the last 20 years, we see a real paradigm shift in the way the service offer has evolved. What was before more a scientific approach with a multi-instrument satellite platform the size of a minibus, are today micro- or even nanosatellites the size of milk cartons which are deployed in constellations of several tens or hundreds of satellites.

There are many providers working on developing new capabilities, and one already operational example of this is Planet, which operates a fleet of mixed size satellites ranging from microsatellites to nanosatellites. In January 2021 we saw the new world record of 143 of such nano satellites launched with a single SpaceX rocket, with Planet alone having 48 satellites on board! It is clear that a wealth of new data is about to start streaming into projects like the GEO.

Volker: In terms of the imaging technology available in space to be used for the GEO service, this really covers a broad range of sensors suited to the respective applications and the specific user requirements.

Optical images for instance provide a kind of “photographic” snapshot from space which can directly be used for analysing the details captured. But it does not stop here, because optical sensors can go well beyond what the human eye can sense, so for example vegetation stress can be detected in the near infrared spectrum which our eyes cannot see as we are limited to the visible spectrum only.

Even beyond that, we are very excited to see more so-called hyperspectral satellite missions, such as ESA’s CHIME (Copernicus Hyperspectral Imaging) Mission becoming available offering hundreds of narrow spectral bands which will greatly expand our capability to support new and enhanced services for sustainable agricultural and biodiversity management, as well as soil property characterisation.

Radar/SAR images on the other hand are also widely used and have the huge advantage to be able to “see” also at night and through clouds, which optical sensors cannot. This can be very crucial in situations of storms and disasters where new data is needed urgently.

What do you see as the main applications of the GEO?

Forbes: For the insurance sector, the applications of the GEO are far reaching. We can automatically ingest risk data, store and monitor it against insured perils, using a huge number of data sources from around the world. Once a trigger event happens, we send that data back to the client system and they act upon it in either their exposure management, claims or other workflows, greatly speeding up these workflows and providing very accurate data from the ground.

For parametric policies, the benefits are multiplied, and we envision far-reaching development and increased relevance of parametric insurance in the months and years to come as a result of initiatives like the GEO.

Nick: The Global Events Observer as it is proposed to us by MIS was to take this EO data and to package it up to present to insurance companies, making it easy for them to assess risk in an age of digital transformation where, driven by digital technologies such as Artificial Intelligence, Simulation, Cloud, Internet of Things (IoT), Big Data Analytics, and AR/VR, the Industry 4.0 is becoming an integral part of Space 4.0.

This actually means that the old fashioned way of sending a person to the site to assess the amount of damages caused by a hurricane for example could be minimised and aided by the analysis performed through machine learning and automatically harvesting and processing all available data sources, one of which is Earth observation data, but not the only one.

Volker: We see also in other projects that it is the right combination of different data sources that provides the best results in terms of information extraction for specific customers, and for the GEO this is the insurance industry, i.e. teams working at every stage of managing risks at insurers.

Another exciting fact is the use of wind data derived from ESA’s Aelous satellite as the service is capable of automatically pulling in important data sources for claims assessments related to damages caused by hurricanes for example.

How are MIS and ESA working together on this project?

Forbes: Shared intelligence is shared success. We are really delighted to have received funding from ESA Space Solutions for the GEO and are grateful for the operational and demonstration support the experts at ESA are providing. Working closely with the team also helps us keep abreast of the latest developments in space technology, ensuring we build a service to customers that delivers value today, but remains fit for the future in this ever-evolving space.

Nick: The European Space Agency is co-funding downstream activities as part of ESA Space Solutions which is the go to place for great business ideas involving space in all areas of society and economy.

Our mission is to support entrepreneurs in Europe in the development of business using satellite applications and space technology and one of these is the GEO project. The rationale behind this project is to really work with potential customers over a time period to develop and test; from a pilot/pre-operational demonstration of the services through to delivery.

Volker: During the course of the project, an ESA Technical Officer actively works with MIS to provide technical and business expertise specifically targeted to the needs of the business idea, like in this case the GEO service proposed by MIS.

On top of this, the growing ESA Space Solutions network offers a unique opportunity to connect and to show successful results as part of this network. All this adds to the credibility and ESA branding which goes with the zero-equity funding ESA provides in supporting innovative business ideas powered by ESA Space Solutions.

How is progress developing to date – what do you plan to achieve in 2021?

Forbes: 2021 is an exciting year for GEO, already we’ve seen the demo progress to build and from May GEO will be put through its paces by players in the London Insurance market via a Beta Program. Following the feedback from this Beta Program, GEO will go live in Q3, offering key insights to MIS clients.

Volker: The demo project started in June last year (2020) and has moved now into the development stage with the aim to start its pilot trials with involved customers at the Lloyd’s market place, so we look forward to this exciting opportunity to see how the enhanced GEO service will look like in operation on a global scale and how it can bring into operation more automated functions which is a prerequisite necessary to efficiently process the vast amount of data.