Index Insurance: Geospatial Intelligence.

INSDEX POWERED BY ARLAS.

21 September 2020, Media Release

London – Toulouse: 21 September 2020 – Skyline Partners and Gisaïa announce their strategic partnership and the launch of, Skyline’s index insurance technology, starting with Solar.

Solar is a unique index insurance solution by Skyline for renewable solar energy, powered by Gisaïa’s geospatial solution ARLAS that will bring a unique risk transfer and management platform for clients to explore, monitor and insure the lack of solar radiation.

The same data used to pay claims will be used to monitor their risks in full transparency for customers and risk carriers.

Combined with faster and guaranteed payouts, Solar will support project developers and green investors in securing better funding terms and help them manage their risk appetite.

“Gisaïa’s expertise in geospatial big data exploration has been proven in the high-tech space industry. Their ability to deliver high quality geospatial visualisation is unprecedented in the insurance industry. We are very excited to partner with them and bring this unique value proposition. Our shared focus on innovation, technology and data cement this partnership for the long term.” Laurent Sabatié and Gethin Jones, Skyline Partners Co-Founders and Directors.

“Skyline Partners have unique expertise and experience in parametric insurance underwriting. Skyline’s ability to keep up with innovative trends and even become early adopters aligns with our quest to explore new spaces to deploy our ARLAS® framework. We are thrilled to find a partner who shares in our push for excellence and positive customer experiences” Laurent Dezou and Sylvain Gaudan, Gisaïa Co-founders and Directors.

Index-based insurance, whilst already a well-known mechanism, has been slow to emerge as an established risk transfer option by risk carriers and customers. This is starting to change and index insurance is now accelerating to complement traditional programmes, in markets where cover gaps or exclusions exist, or conditions are tightening.

With Skyline aims at facilitating and spreading the use of index insurance, as an efficient and transparent means of risk transfer with fast and guaranteed access to cash at the occurrence of adverse events.

New upgrades will soon be released, to accompany existing Skyline index solutions for renewable wind energy, agriculture and natural catastrophes. In addition, Skyline has also developed a beta version of index insurance technology for distributors and risk carriers that integrates the same visualisation capabilities, but also index risk modelling, large scale pricing and payout trigger notifications.

By combining Skyline’s index modelling capabilities with Gisaïa’s big data geospatial expertise, the partnership will create a powerful and unprecedented proposition for the insurance industry. It will bring significant value to customers, brokers and carriers to help them comprehend, manage and transfer risks, in line with their specific needs.

For more details

Contact Skyline to discuss index-based insurance: enquiries@skyline.partners

Additional RESOURCES

Please follow Skyline Partners and Gisaïa for updates on the joint venture:

Follow Skyline Partners on Twitter, LinkedIn

Follow Gisaïa on Twitter, LinkedIn

About Skyline Partners

Skyline Partners is a private-equity funded UK insurtech company, focusing solely on index-based parametric insurance. Our solutions are data, technology-driven, and underwritten by top tier international re/insurers with superior ratings. Skyline has developed its own index insurance technology platform and aggregates high-value data from multiple sources. Our initial focus is on Renewable Energy, Agriculture, and Natural Catastrophes, but our ambition is to develop index solutions for all classes and segments. For more information on our vision and solutions, visit www.skyline.partners

About Gisaia

Gisaïa is an expert in geospatial intelligence with experience in facilitating rapid deployment of technology solutions for deeper geospatial-big data analysis. Our ARLAS ® framework is built to serve efficiently: every function is developed to not only support the ease of operations but also, deliver expected results. We have tried and tested the potential of ARLAS ® framework’s versatility in diverse use-cases, with great results. For more information about Gisaia visit: www.gisaia.com and ARLAS: www.arlas.io

Watch video demo here:

TRACING THE DANGER IN THE AIR

If we understand it, we can fix it

July 27, 2020

4.2 million people die every year.

That number of deaths is linked to outdoor air pollution as recorded by the World Health Organisation (WHO). When you include indoor pollution, the number goes up to 7 million people every year.

While most of these deaths are in developing and middle-income countries, major cities in developed countries still record higher than recommended levels of substances linked to dangerous pollutants.

So, what are these pollutants and how do they affect people?

The OpenAQ database that is used in ARLAS’s – also an open-source framework- demonstration, collects data from 12,000 stations spread across 93 countries. The stations monitored by OpenAQ record diverse air pollutants; NO₂, SO₂, NO₃, BC, PM₂.₅, PM₁₀, and CO₂.

NO₂, Nitrogen Dioxide, is produced in abnormal quantities by human activities like burning of fuel, motor vehicle emissions, and power plants. It is toxic and easy to inhale, known to trigger asthma.

The PM₂.₅ is also a killer pollutant. Simply referred to as ‘Particulate Matter’, it is composed of solid particles and liquid droplets. It is highly inhalable causing many health problems, especially linked to the heart.

PM₁₀ is four times bigger than PM₂.₅ but just as dangerous because it can also be easily inhaled.

The trouble with air pollution is that it affects even those who didn’t create it. The big question would then be, how to trace the danger in the air and prevent exposure to these deadly agents? While many answers speak to long-term strategies, the immediate response involves aggressive monitoring and warning systems that give individuals the ability to make decisions that protect them.

Having reliable air pollution monitoring tools like ARLAS ensures that everyone in the decision-making chain is well equipped to act.

THE COVID-19 SCENARIO

The world ground to a halt and immediately opened the skies – quite literary in some locations – to the possibilities of tackling air pollution. The near-global lockdown courtesy of COVID-19 made it apparent for ‘pollution effects believers and non-believers’, that the Earth had been choking.

On the left, Beijing on a clearer day and on the right, the city covered in smog, a feature common with high pollution

It was no surprise to see pollution levels go down as mobility reduced and non-essential factories’ tracks go on a standstill, cutting pollutants by significant numbers.

China for example is on record as the highest contributor to CO₂ levels globally. During their lockdown, their CO₂ emissions went down by about 25%, equivalent to 150,000 tons.

To put that into perspective, that is nearly five times what Angola emits yearly.

We tested the theory of human contribution to air pollution. Using ARLAS, we compared March – April 2019, March – April 2020, and June 2020 data. This period represents, before and during COVID-19 for the same periods and just after the lockdown was lifted in some countries.

Comparision of the Evolution of PM₂.₅ emissions in the Beijing/Tianjin/Hebei/Shandong provinces from January to March 2019 versus 2020

Focusing on provinces around Beijing, ARLAS shows PM₂.₅ levels reducing by about 30% over the period when the lockdown is ongoing. It is clear from this demonstration that reduced human activity also reduced the volume of pollutants released in the air. There is also a significant reduction in PM₁₀ levels.

Some activities that are traditionally low contributors to pollution like home heating and data centres increased their pollution contribution as human activities migrated home and online. This kept some pollutants volumes high.

OTHER ARLAS GENERAL OBSERVATIONS

On ARLAS, a dive down to specific stations reveals a more localized picture.

Based on the dangers associated with PM₂.₅ and PM₁₀, the WHO advises that exposure to these two agents should be kept below 20μg/m³ for PM₁₀ and 10 μg/m³ for PM₂.₅. Some locations record levels that are over five times higher than the recommended values.

Residents of New Delhi benefited from a 56% drop in PM₂.₅ emissions between 26 April – 7 May, 2020, compared to the same period in 2019.

But, it was still, way above the recommended threshold.

Comparision of the PM₂.₅ emissions over New Dehli city between 26 April – 7 May 2019 (mean of 95.8μg/m3) and 26 April – 7 May 2020 (mean of 43.4 μg/m3)

NO₂ emissions also have an adverse effect on health especially when higher than 100μg/m3 on an hourly average. Between January and August 2019, Paris experienced levels above 100μg/m3 85 times. In 2020, it was cut down 21 times to only 4 measures of similar quantities that were recorded at the end of June, just after the lockdown.

Temporal comparision of the 1st Semester 2019 versus the 1st Semester 2020 of NO2 emissions higher than 100μg/m³ over the City of Paris

Zoom on the legend Over the City of Paris :
Above: Distribution of numbers of measures of quantities of NO₂ emissions. Here, you see a selection of emissions above 100μg/m³.
Below: Distribution of the average of NO₂ emissions over time (In the 1st semester of 2019, there are 13 peaks. For the same period in 2020, only 1 peak).

AIR POLLUTION STAKEHOLDERS

Strict monitoring of air pollution is a growing phenomenon with more stations being set up to facilitate this. The European Union region has stringent policies on pollution which have resulted in lower deaths linked to air pollution. France for instance has reduced air pollution linked deaths by half in only 25 years.

Pollution data gives the ability to analyse it and fix it.

An ecosystem of monitoring and prevention would ideally save many lives: different players acting from different angles, easily and quickly guided by data.

To start, policymakers like scientists and regulators, require needlepoint precision to determine their actions.

Travel companies that issue alerts on destinations based on pollution levels could help many at-risk groups to make informed choices – for example, people with preexisting pulmonary conditions. Could this lead to cheaper insurance premiums for travellers to/during low pollution zones and periods? Or even better, saving lives.

Individuals may also want to proactively stay connected to pollution information within their localities or elsewhere they intend to visit. Access to mobile applications that have credible up to date data would put pollution-fighting power in their hands like; where to find a job or buy a home amongst other human habitation decisions.

WHO is keen to have more stations set-up because the more air pollution data is collected, the clearer the picture. The OpenAQ currently collects up to 588 million air quality measurements. This is big data that ARLAS can quickly and easily process for Air pollution analysis.

AN OPEN APPROACH TO ADDRESSING AIR POLLUTION

The irony of the COVID-19 pandemic is that within the initial limited time of confinement – mid March – end of May 2020 – more lives may have been saved from the direct effects of exposure to pollution than were lost due to COVID-19 during the same period.

Scientists are already linking higher critical COVID-19 impact amongst people who were previously exposed to high levels of pollution.

All efforts towards air pollution monitoring are expected to help reduce premature deaths linked to toxic air. If you are working in monitoring, policy setting or warning mechanisms around pollution, check out our demo to get a glimpse of Arlas at work on OpenAQ pollution data.

Get in touch with us at contact@gisaia.com if you would like to discuss tools and services to get you acting faster.

EXTERNAL SOURCES

Open AQ Website

ESA Article “Global Air pollution maps now available”

WHO Article “9 out of 10 people worldwide breathe polluted air, but more countries are taking action”

Forbes Article “Coronavirus Lockdown May Save More Lives By Preventing Pollution Than By Preventing Infection”

Our world in data Aricle “Air pollution is one of the world’s leading risk factors for death”

EXPLORE EARTH OBSERVATION DATA WITH ARLAS

June 30, 2020

Geospatial Intelligence from Earth Observation

Today the United Office for Outer Space Affairs (UNOOSA) records over 2000 active satellites are orbiting Earth. Most are earmarked for other functions like communications, technology development, navigation and space science. 884 of these satellites are deployed for Earth observation.They are observing Earth and sending back diverse information from; biological, physical and even chemical data on our planet.

The interest in Earth observation started with the launch of Sputnik back in 1957 and grew exponentially in the last 20 years. Earth observation has since evolved to take advantage of other remote-sensing technologies to include extra dimensions to the spatial and temporal values collected, like hyperspectral images, creating big linked geospatial data.

EXPLORING THE RICH MATTER IN EO BIG DATA

A study published by Acta Astronautica, observes that while the initial players in Earth Observation (EO) were governments and large corporations, the field is witnessing an increase in the private sector actors. This is not only ramping up the volume and velocity of the data collection but, it has made EO data more accessible and affordable.

10 years ago fewer active satellites were observing the earth. Terabytes of data were generated. Yet, we had no capacity to neither store nor analyse that volume of data. Using the right technology, institutions can now access the full data value-chain and mine gems from data embedded in earth observation products even with current greater volumes and higher velocity.

The increase in EO producers – satellite manufacturers and launchers – is in tandem with EO consumers – data processors and analysts. Establishments find newer use-cases for EO objects that are especially valuable in geospatial intelligence: from environmental analysis that informs policies to urban planning that leads to smarter cities and not to forget IoT tracking.

Firms are especially investing significantly in big data analysis for business intelligence. Whether the intent is to reduce costs or save time, the outcome in speedy decision making leads to increased revenue.

Gisaia’s ARLAS® framework, an open-source technology, is developed to create high-value information for policymakers in the public and private sectors. Currently, it is deployed by France’s consortium of Earth Observation and Environmental sciences, Theia and also by Data-terra as Dinamis. They both use ARLAS technology for Earth observation cataloguing and analysis of Earth observation objects.

Example of filters using cloud coverage and location by name on Dinamis

Search results based on query on cloud/time/location on Dinamis over La Reunion

SOME PRICKLY POINTS WITH EARTH OBSERVATION ANALYSIS TOOLS

If we look back, we see that the journey for the pioneers in Earth observation was not always easy. Billions of earth observation objects were left in the archive with little to no exploration due to the amount of time, and expertise that was required to sift through them.

Even with improved technology, many EO producers find it difficult to aggregate data from different sources for studies. Like, how do you know that you have chosen the right image to analyse? This alone, can be a tedious exercise that takes a lot of man-hours, sometimes, looking at millions of images with no certainty in results.

While some tools can now visualise the objects, often there isn’t an option to pre-visualise, allowing one to check if the image meets a criteria for the query like; cloud coverage or mission type.

It is no wonder that most of these tasks are mostly relegated to data scientists and IT professionals within teams as they often require some knowledge in coding.

ARLAS® FOR EARTH OBSERVATION

As the ARLAS® framework was initially conceived for earth observation, its development has evolved to absorb some prickly points in geospatial big data technologies’ usage.

The main principle behind ARLAS is to make it accessible. One does not have to be a coder nor a data scientist to use it. This then cuts the amount of time spent on the technical details – setting up and filtering – to be used for instantaneous query of the data.

One of the many resourceful features of ARLAS when conducting a search is, getting results to your query plus, suggestions close to your query. This gives you the option to quickly pick the right image.

There are several built-in time-saving configurations like the ability to save your filters if you use them often. Advance filters by detail like excluding images with cloud coverage or even sharing your URL with teammates to quickly view before making decisions.

Example of Catalogue with Theia over South of France

If you work with data from different sources, ARLAS is also able to aggregate them under one platform on a user-friendly interface. Because it is open-source software, there is no vendor lock-in giving you;

Freedom to adapt it
Freedom to extend it
Freedom to change it

CONTACT US: OPPORTUNITIES ABOUND

There are numerous prospects in earth observation objects analysis and we have expertise in geospatial intelligence. If you have questions about how to combine the volume, velocity and variety from EO to create high-value information, get in touch with us at contact@gisaia.com

Automatic exploration and classification of behaviours of migratory birds with ARLAS

April 09, 2020

To better understand animal behaviour, many researchers are equipping some individuals with GPS beacons and sensors to track their movements. This new data are enabling major advances in the understanding of our biodiversity but also of climate change. To make it possible to analyse data that are voluminous, it is necessary to use high-performance tools.

At Gisaïa, we are sensitive to environmental concerns and are developing ARLAS Exploration, an open-source software for geo-spatial data mining capable of responding to Big Data challenges. We have therefore decided to put ARLAS to the test in order to make it possible to explore animal data interactively.

Data from a study on a storks population was used, and we will see how ARLAS can be used to discover the behaviour of these migratory birds. We will also see how it can accompany the implementation of Machine Learning algorithms to automatically detect some of these behaviours.

A data set of storks

THE DATA SOURCE

The Max Planck Institute of Animal Behavior, a German research institute that studies wildlife, recently launched Movebank, a platform that aims to centralise data from animal studies conducted by researchers around the world, to encourage collaboration and make it possible to cross-reference data from different studies. This initiative also aims to make this data freely available to the general public.

In a study by Cheng & al. (2019) [1], researchers equipped a population of 169 storks with GPS beacons and collected data between 2013 and 2019. These data were retrieved from the Movebank [2] platform as CSV files.

The white stork

The white stork (Ciconia ciconia) is a large species of wading bird in the Ciconiidae family. Its plumage is mainly white, with black on the wings. This species has been the subject of protection and reintroduction programmes and is mainly found in Eastern and Western Europe. The stork is highly migratory and winters in Africa, making its movements particularly fascinating to study.

Processing the data

The data was first processed in order to group and link together the successive observations of the same bird and to calculate the travelled distances. This type of processing is done using ARLAS PROC/ML, our massive distributed processing platform. The data thus formatted is then integrated into ARLAS Exploration and our storks are then ready to be explored.

ARLAS, a fluid and interactive exploration

ARLAS Exploration is a map-centric application that allows to appreciate the spatial dispersion of the data. A bar of graphs on the left of the application also allows to visualise and filter the other dimensions of the data, so we can observe the distributions of different quantities such as the travelled distance, but also the altitude and the speed. At the bottom, the timeline allows to see the temporal distribution of the measurements made.

SOME CLEAR OBSERVATIONS

It is clear that the traced storks move within a perimeter that extends from southern Germany to West Africa. There is a peak in the number of observations in August 2014. In addition, data are actually available for 81 birds for a total of about 7 million positions.

EASY EXPLORATION ON ARLAS

The various graphs allow you to filter the data on the dimensions represented and the whole application is instantly updated with each new selection. It is also possible to navigate the map on certain areas and filter according to geographical selections drawn with a tool on the right side of the application.

Depending on the amount of data to be displayed in the application window, ARLAS switches from an “aggregate” mode, density maps ideal for general visualisation of the flows, to a “features” mode, the detail of the actual data to observe the actual paths of the storks. We can thus isolate interesting behaviours, and see that some storks seem to use hot airstreams to gain altitude, for example:

The actual route displayed can be colored in different ways, depending on the speed or the bird ID for example:

These two representations allow us to understand that the four selected storks named Hans, Schwitza, Kiki et Julia, move together at the same pace over this period, while still being able to clearly distinguish the four storks.

Thus, ARLAS Exploration is a tool particularly adapted to interact intuitively and interactively with bird positioning data, even when the volume of data becomes large. This makes it a strong partner for researchers.

Towards the detection of migration

The observed storks tend to migrate great distances to change habitat locations. One can then distinguish two attitudes: staying in the same area (local) or travelling to change areas (travel). Both of these behaviours are locally visible to the naked eye in ARLAS Exploration, but automatic detection of these activities could make it possible to study the migrations of all birds on a large scale in an extremely efficient manner. This is why we have chosen to use Machine Learning algorithms to automate this identification. Supervised learning was carried out to train the classification model.

This process was therefore carried out in several stages:

Construction of a training set
Calculation of new indicators
Choice of the classification model
Viewing the results

Construction of a training set

A supervised classification model needs training data to learn how to recognise targeted behaviours. In our case, it is necessary to annotate our data. Each fraction of a trip must be identified as ‘travel’ or ‘local’.

ARLAS Exploration is a tool particularly adapted to the creation of training sets since it allows to assign a label to the current data selection. It is therefore possible to manually identify the parts of trajectories corresponding to a large displacement (“travel”) or a local activity (“local”) and label them as such.

In practice, 4 birds have been labelled as such:

For each bird, a period of approximately a year was used to capture at least one round trip in the migration. We therefore have a total of 334,372 fragments (interval between two measurements) that will be usable for training the model.

Calculation of new indicators

The quality of a Machine Learning algorithm depends above all on the quality of its training data. Once a sufficient number of representative fragments of the behaviors to be detected have been labelled, it is necessary to choose the sizes that will be given as input of the model. First of all, the tagged data can be retrieved using an ARLAS API, available in Python among others, which allows the training data to be downloaded.

In our case, the selected features illustrating the movement of these birds will be based on travelled distances and “as the crow flies” distances” calculated over different time windows. These features are not present in the raw data and can be calculated using ARLAS PROC/ML, our processing platform adapted to large volumes of data.

Once these new quantities have been calculated, they can be used to train the different chosen Machine Learning models.

Choice of the classification model

Several classification models have been tested. In order to be able to compare the quality of these classifiers, the calculation of indicators is necessary. Since the classes are disproportionate (~6% travel in the training set), several metrics were used to correctly evaluate the quality of ‘travel’ detection. A “cross-validation” is performed to avoid overfitting by partitioning the training set and measuring classification performance on data that is not used during the training of the model.

The metrics used are based on the confusion matrix of the prediction:

We have:

Accuracy: Overall proportion of good classification
Recall: Share of correctly detected real ‘travel’
Precision: Proportion of detected ‘travel’ that is actually real
Specificity: Share of correctly detected real ‘local’
F1-score: Harmonic mean (trade-off) between Recall and Accuracy

As all experiments are performed under the same conditions, the models can be compared with each other, in particular thanks to the MLFlow tool used to record the results. Finally, after numerous experiments, an XGBoost classifier was chosen, both for its performance and its training speed.

Viewing the results

Once the model has been chosen and trained, it can be applied to other birds and the results of this migration detection can be exported to ARLAS Exploration thanks to the tagging system (also available via API). It is then possible to visualise the results directly in the application. This allows a better understanding of our model by quickly identifying on which part of the data the predictions would fail. It is also possible to validate or correct the results, which makes it possible to increase the training set and to train the model again on more data. In the case of our storks, the model was applied to 26 birds, corresponding to 2,650,000 fragments.

We can also follow the track of a particular bird and identify the different stopping places along its route. For the stork named Zozu, for example, the following results are obtained:

We can also date the great movements of these storks. If we consider the predicted travel fragments of the stork named Zozu, for example, we can observe the different peaks on the timeline, which makes it possible to identify the periods of the year when the bird migrated:

If we look at one of these peaks in particular, we can date these migrations very precisely and see the different stages of the journey, so this journey between Switzerland and northern Spain between 21/08/2015 and 29/08/2015 took place in 7 stages:

The migration of these birds can also be explored on a larger scale. If one selects all the fragments in ‘travel’ for all the labelled birds, one can see a migration corridor following the Mediterranean coast towards Spain in the south of France:

Finally, it is possible to identify the different living places favoured by the storks during their journey around the Strait of Gibraltar by selecting the ‘local’ predicted fragments for all the storks:

The automation of migration detection has therefore greatly facilitated its analysis and ornithological experts can now focus on the variations in dates and destinations of the migrations undertaken by the storks. The possibility of cross-referencing this information with other data, such as meteorological data, can provide an even better understanding of the behaviour of these large migratory birds in relation to, for example, climate changes.

Conclusion

We have seen that ARLAS is particularly well suited for exploring position data of storks. The interactive map navigation provides valuable information on the behaviour of these migratory birds. But ARLAS can also be used to support the production of Machine Learning algorithms by facilitating the creation of training set and the visualisation of classification results. Finally, once the Machine Learning models have been trained, it is possible to apply them to large-scale data with ARLAS PROC/ML and see the results in ARLAS Exploration. All these results are available in a demonstration that is available at demo.

If we have focused here on migration, many other animal behaviour could be the subject of such studies. Obviously, ARLAS can be applied to all kind of geo-traced animals, but also to all geo-referenced data. Feel free to have a look at the other application examples on demo.arlas.io.

Thanks:

We would like to thank the director of the Max Planck Institute of Animal Behavior, Dr M. Wikelski ,and his team for providing this data.

References:

[1] Cheng Y, Fiedler W, Wikelski M, Flack A (2019) “Closer-to-home” strategy benefits juvenile survival in a long-distance migratory bird. Ecology and Evolution. doi:10.1002/ece3.5395

[2] Fiedler W, Flack A, Schäfle W, Keeves B, Quetting M, Eid B, Schmid H, Wikelski M (2019) Data from: Study “LifeTrack White Stork SW Germany” (2013-2019). Movebank Data Repository. doi:10.5441/001/1.ck04mn78