Detecting and mapping COVID-19 across Luxembourg

Published on 18/06/2020

A major part of the battle against the onset of the COVID-19 pandemic is detection and mapping and that’s exactly the focus at LIST of “Towards an Integrated Geospatial pandemic Response system” or TIGER project for short.

WHAT IS TIGER RESEARCH ABOUT?

“The idea is to develop a web platform to help policy and decision makers in crisis situations such as the current, pandemic to make better decisions,” explains Ulrich Leopold TIGER project’s Principal investigator. “We have already developed this platform at LIST, but for other purposes, plus we have developed tools that can handle the spread of diseases in space and time,” he added.

Indeed, the TIGER project uses a platform and its tools, which has been developed by LIST over the past 10 years for completely different reasons such as rainfall prediction, air quality or soil pollution prediction, “but since we have some similarity with epidemiological data, we can apply these tools as well and with this give a full picture,” Ulrich stated. “We currently don’t know where exactly - not down to neighbourhoods, not even down to municipalities - where infections are happening, so with such an approach we will obtain a better image of those patterns.”

Once complete the TIGER project will be a major contributor to Luxembourg’s COVID-19 TaskForce and having the flexibility of being implemented into an interoperable geospatial web platform, which in turn, could be linked to - or integrated into - current and future initiatives on the spreading and control of infectious diseases.

The idea of the platform is similar to street mapping but instead of mapping local services, it will display and locate virus infections. With a very high resolution, it could focus right down to individual buildings, but Ulrich pointed out, “this has some privacy issues. I don’t know yet if we will be able to get that high resolution information, but let’s assume we do. We can maybe then give an aggregated overview of neighbourhoods, districts of cities, or at a municipality level. But it also depends a little bit on decision makers.”

With the current controversy and scrutiny of COVID-19 tracking apps, Ulrich was quick to point out that the TIGER project is not at all similar. “We are not going on to mobile phones, it is just when people have registered their address with health services, so we take that information, or there might also be a survey coming from LISER institute, where they ask users for some self-reporting, and we can use these data.”

The TIGER project has major impacts and innovation potential due to its integrated approach combining distributed data, methods and tools to provide space and time evidence for multiple experts, researchers and stakeholders, to take improved decision for lockdown measures leading to a controlled reopening of social life.

The project uses probability algorithms for a replicable, reproducible and scalable response system that will help reactions more rapidly in future outbreaks of infectious diseases and can provide better coordination with neighbouring countries.

“When we make predictions, we get one map where it would give you an average situation, but since this is a probability approach, we can run many simulations. For example, we can run 1000 simulations for Luxembourg where we get 1000 maps of potential distribution patterns. And if we take an average of these 1000 maps we get an average prediction map. With this we can give a range and even show prediction errors,” Ulrich explained before going on to state that “with this average prediction we can show another map that just shows error-associated predictions; maybe in one area we are more confident than another where there is less data which then gives us an idea where we should sample more".

Ulrich clarified that LIST’s TIGER project is not just a report generator based on calculations for stakeholders. “You will be able to go onto a platform and interactively query the data with a login access. Also, the platform has an advantage as you can link in other data from, for example, the Luxembourg Geoportal population information. Theoretically we can also link it out because it follows some certain standards for data exchange.” Results can be easily shared through interoperable web services to the different expert teams and the government in forms of maps, tables, graphs and summary statistics at different aggregation levels.

The end users the platform is currently aimed are the likes of Luxembourg’s decision makers, other researchers with whom the TIGER project can exchange data, the country’s COVID-19 TaskForce, other epidemiologists, “and maybe eventually the public, but this has to be decided by decision makers,” concluded Ulrich, “it is not up to us to make that decision”.

WHAT IS LIST SPECIFIC CONTRIBUTION?

“TIGER is entirely a LIST project. However, the specific contribution is the web platform that is already in existence, developed over the past 10 years that we have adapted to disease data. Then there’s the core contribution with what we call spatial-temporal models or algorithm so you can actually map disease outbreak patterns across the country and you can also map this over time.”

WHAT IS YOUR SPECIFIC ROLE?

I am the Principal Investigator so I manage the project. I put out the challenge and the vision and I help a bit implementing these algorithms, but others mainly do that on the team. Then some of the teamwork on data provision making all data ready so that it can be used uniformly. One person implements those algorithms and two people further adapt the web platform.

ABOUT ULRICH LEOPOLD:

Ulrich Leopold received an M.Sc. in Physical Geography and GeoSciences at Trier University and did PhD research at the University of Amsterdam in Spatial Uncertainty Propagation Analysis. He is a Senior Research Associate in Geocomputation at the Luxembourg Institute of Science and Technology (LIST) and has 20 years of experience in geospatial analysis and modelling, geospatial software technologies, geostatistics and spatio-temporal uncertainty analysis in national and international projects. Ulrich has been leading various national and international projects and developments in the field interoperable web-based geospatial software platforms, geostatistical applications, geospatial uncertainty propagation analysis in various (environmental) application fields, such as Water, Air, Soil, Logistics, Renewable Energy, Sustainable Urban Planning, Noise, Earth Observation, Smart Cities. He has been securing funding for numerous research and development projects by FNR, INTERREG NWE, H2020, ITN Marie Curie, ESA as well as Industry.

Team and competences

The team has 10-20 years of experience in the field, forming an interdisciplinary cross-departmental team at LIST, providing competences in GeoCopmutation, Environmental Modelling, and Computer Science. This team has already worked in previous research and development projects together where team members have developed parts of the technologies and methods to be built upon in the TIGER project.

Ulrich Leopold (PI), Christian Braun geospatial analysis, Arturo Torres spatio-temporal probability model and simulations, Sukriti Bhattacharya - distributed computing, Philippe Pinheiro interoperable web platform architecture and developments, Mickaël Stefas Web client frontend