This scientist explains how solutions like GeoHealth are very relevant to help reduce risks during the end of the quarantine. She reminds people that MNOs' operating data are a very reliable and useful source to understand the circulation of the virus. And, without any particular app. deployment.
« These data remain anonymous and aggregated: one cannot track the users' routes. The aim is not to individually control people's movements, but to analyze population transfers and anticipate the consequences on the epidemic.» she says.
In France, the end of the quarantine will be adapted by region and will be coupled with a ban on traveling more than 100 km (except in specific cases). Vittoria Colizza relays one of the use cases that we are proposing with GeoHealth: verifying the collective respect for government measures: "If we combine this information with spatial diffusion models, we can not only visualize the epidemic situation but also anticipate. We can measure how transfers between two regions facilitate the spread of the virus and thus adapt hospital needs.”
Translation of the article published on La Tribune on May 4th (in French).
While the examples from Singapore and Israel show that tracking apps like StopCovid are ineffective, is the use of technology a wasted effort in the fight against Covid-19? No, there are other solutions that France should consider. By Yann Chevalier, CEO of Intersec, and Jérôme Faul, managing partner at Innovacom.
As the date of May 11th approaches, the organization of the end of quarantine raises every day its share of difficulties and challenges to be met. One of these priorities is the “Stop Covid” application, on which the government is working hard. Designed to allow end of quarantine in the best conditions, "Stop Covid" can be downloaded by each of us on our mobile, on a voluntary basis and anonymously. But if the government seems to have won the battle of opinion on the use of this application, it still has to demonstrate that its use will save lives and to specify which strategy the application will serve as recalled by the CNIL in its opinion given Sunday morning.
Unfortunately, it is estimated that the Singaporean TraceTogether app. only highlights 3% to 5% of contamination from an infected patient to a healthy patient. Due to a poor download rate (one sixth of the population), only made for holders of a smartphone (77% of the population approximately in France); and to a psychological obstacle to declare themselves sick on the application. As people affected are not sure how the information about their condition will be used in the prevention strategy, or even wonder about the consequences (in terms of access to public transports, taking out a mortgage, etc.), one can easily imagine that the efficiency rate drops to 1% or 2% in a less disciplined context like in France.
Israel, which was carefully monitored at the start of the epidemic and which has made a comparable application its spearhead in the fight against the virus, is now under pressure from public opinion. TV shows come up with new anecdotes every day that show the aberrant limits of the system: pizza deliverers see their application spot ten resident phones each time they enter a condominium. Without ever being in contact with the occupants, if one of the occupants is sick, the delivery man himself is considered as sick by the app. and his application then "contaminates" those of other residents. In a country that uses technology in a restrictive way, this pushes a large fraction of the population to an unjustified quarantine.
Considering these misadventures, some governments give up. The Netherlands have just given up on the development of the application for technical issues. It takes the cooperation of Google (for the Android system) and Apple (for iOS) for the Bluetooth data to reach the application permanently. This requires an update of the entire fleet of smartphones. The two American giants propose to create a native system in their OS which will create a database of frequented contacts. Then each smartphone user will be free to download a governmental application which will make use of it, or not ... The potential monopolization of our private relations in the physical world by the GAFA is an additional red line to which certain European governments begin to oppose a veto.
Is using technology to help with this crisis a waste of time? No, and paradoxically we should turn to the countries that have the best control of the epidemic to find the solutions. South Korea, in particular, a democratic country in Southeast Asia, has developed a strategy based on a health pillar and a technological pillar. The health pillar is elementary: systematic use of masks, profuse hydroalcoholic gel, telework, social distancing, restriction of movement, and above all systematic and large-scale use of screening tests in order to break the chains of contagion. The technological pillar is also instructive.
Location data is actually used, but not to trace contacts (it is technologically impossible on the basis of geolocation whatever the source). A weather forecast for the circulation of the virus is compiled and updated by the authorities. The location data of the infected people are extracted every day by the telecommunications operators in order to reconstruct their routes in a completely anonymized way. Each district thus receives an evolving score, according to its frequentation by sick people during the last two weeks. The health authorities may map the whole country, both thanks to information calculated automatically by algorithms but also to additions and corrections done by humans.
Thanks to this public knowledge database established by the State using data collected from patients and immediately anonymized (we never keep the identity of the location points noted), a "virus weather report" is made available to citizens. Anyone can consult it via an app. or can subscribe to an SMS notification. It thus becomes easier to adapt your degree of quarantine to exposure to the virus in your neighborhood. Likewise, authorities identify new clusters at a very early stage and can decide on quarantine measures, curfews, or administrative closings of shops or public places.
South Korea is currently reporting fewer than 20 new cases a day and the death toll has remained below 300 in total since the start of the pandemic. So why don't we put similar health and technological resources within the European Union? Our telecommunications operators have recently shown that they have the technology to conduct anonymous surveys. It would be advisable to seize it and to exploit them in a systematic way to come out at best and as quickly as possible of this epidemic which we have already taken too long to try to stem.
The magazine "Décideurs" (Decision-Makers) has just released an interview of Yann Chevalier.
Here is the translation:
Décideurs: Tell us how location data are used.
Yann Chevalier: When we move, our phone keeps interacting with the network and this latest keeps a trace of our approximate location. MNOs need this location to run the service. These data may then be monetized. Anonymized, they are processed within studies to better target who to send promotional offers to but also to give insights to public authorities such as a city, a region... For example, if you consider the future CDG Express train line in Paris Region, its layout has been defined thanks to SFR’s data, using the Intersec Technology. Populations’ movements have been analyzed: where are people coming from? Where are they going? What is the frequency of their journey? Etc. Understanding their movements was key to build the right path, especially given that this new line will cost around €1.4B.
What exactly do you do as a company?
We are a software editor. Our software processes technical and commercial data. We develop our solutions and install them at our customers’ premises – historically, telecom operators.
These solutions allow to understand customer experience and to personalize it, to model population flows, to spot abnormal behaviors, etc.
What do you propose to help fight the pandemic?
We have gathered some features that were already developed to build GeoHealth. GeoHealth is a solution allowing public authorities to spot contamination clusters or to warn people in a zone or about to enter it. Police missions could be eased thanks to anonymized data: once processed, they could nurture a heatmap in real-time to see where all of a sudden a group of many people appears. It could be used to make sure beaches are empty, and to send alerts for police to go and to scatter the crowd, or to send SMS to the people in the area to warn them about the danger of gathering. We also know when a foreign SIMcard is activated in France. Here, once more, one can easily see how location can help understand the pandemic evolution and understand the links between the risk zones and the entries on our territory.
How about privacy with this solution?
First of all, before being processed, data are anonymized.
Second of all, the location is not so precise, it is about a 300m-wide radius. This precision allows us to know if too many people entered a mall but we would not be able to get the insights for a defined shop within the mall. GeoHealth also enables to inform people so as to prevent issues. It is not at all a tool made to understand who does what and to punish. We only count people in zones.
However, your votre solution seems to be more successful outside France and even Europe?
This is a subject we are currently dealing with mainly for South-East Asia, Northern Africa, Middle-East and Canada. Two reasons to that: First of all, Asian countries were the first ones to see the pandemic. Local Public authorities have been looking for technologies to deal with the virus and its consequences sooner than Europeans. Other countries, like Canada for instance, where the virus got more recently, generally use technology with more parsimony. And yet, in order to save lives, they do not hesitate to innovate, put solutions in place and be ready to step back if the solutions are misused in the end. In France, country where the respect of personal liberties is monitored very closely, the approach is the opposite: making sure the solution we launch makes the case for the issue beforehand.
In France, an SMS sent to people to explain the quarantine has been widely discussed. What are your thoughts about this?
The government asked operators to send an SMS to their subscribers’ bases to announce the quarantine. It was a message of general interest. One should not lumps everything together. Yes, some solutions can be very intrusives, for instance in China, a country tracking its populations’ personal data shamelessly. But that’s not the case for all technological solutions. Take French Tier-One Orange. Their data have been analyzed to give authorities an idea of the rate of people living Paris city when the quarantine started. Here, there is no issue: data are and remain anonymous – by the way, the Cnil (French National Commission for Data Protection and Liberties) does not rule these calculations since they do not provide personal informations. We can understand which city, which district are even neighborhood people come from but that’s it. Idea is to understand the flows to adapt hospital services accordingly.
Are these solutions expensive?
If these kind of solutions where used to the maximum of their potential by let’s say the French government to manage the epidemics, total budget would be less than €5M.
While Covid-19 is spreading all over the earth, mobile carriers would definitely help public authorities and populations by getting the GeoHealth solution from Intersec, a French scale-up that has been working on location-based insights and mobile messaging for over a decade.
As we all know, the virus is extremely contagious. Some countries are still at the early stages of the epidemic, mainly in Africa – Senegal, Mali, Mauritania, Angola, Namibia, etc. For these countries, being able to review the complete paths of the very first infected individuals in a region is key to nip the epidemic in the bud.
That is exactly what the GeoHealth solution does. It can help sanitary agencies see the last known positions of an infected person up to 15 days back in time, enabling them to get accurate information about real stops of the individual, showcasing places where actual virus contamination may have occurred, so that investigations, communications and field actions are optimized.
Moreover, authorities need to communicate adequate precautionary messages to people who were previously located in the hazardous zones and placemarks that the infected person was in, in the last 2 weeks. These people need to know they might be at risk and get instructions to be examined by a doctor as soon as possible. Once again, GeoHealth can do this as the tool not only enables authorities to spot risk cases, it also enables them to send messages to warn all people who are / were in, or about to enter, a defined area.
Imagine in France for example, if the people attending the evangelic meeting in Mulhouse had been warned by such a tool, they would have followed the quarantine rules and contaminated way less people.
This GeoHealth solution proves very useful since it allows authorities to contact people who for example attended to shows or concerts in a given region during the last 5 days, where and when contamination and death rates have dramatically spiked up, inviting them to follow very precautious protection measures and stay in quarantine for the coming weeks. And, before the epidemics has reached its peak and spread all over the country - authorities may need to get global and regional statuses on the location of infected people. This would help identify new major clusters, and then size and delegate concrete field actions, in regions or provinces.
For countries where quarantine is already in place however, it is important that national and local authorities may convey the adequate instructions at municipal/inter-municipal or canton/county level. Sending the proper message to the nearby populations, informing them about new rules to follow or covered market opening hours is important and location-based messages are the best way to ensure that the population reached by these messages is in the right neighborhood.
It is also fundamental for authorities to be able to inform populations of protection instructions for application on a cluster-basis (e.g. fever monitoring, containment decreed, etc.). GeoHealth, would definitely be helpful for many mayors in these difficult moments. If only they could turn to their MNO to get this service!
Going further, this GeoHealth solution could be the backbone of a vertical solution, which would consist in gathering individual infection and healing reports from medical crews, updating the databases with fresh statuses as they are entered, and visualizing the situation thanks on map-based tools.
It is quite easy to put in place and could definitely be of great help for the sanitary authorities.
Let’s put together an example scenario to better understand the potential of this solution:
Let’s say, the epidemic has come to Canada and the authorities (both national and regional) must generate and share daily reports on the accurate live epidemic propagation situation. GeoHealth can provide two different visualization experiences to provide two levels of situation understanding. Firstly, a global country-wide view featuring a heat map; secondly, a projection of the infection cases by administrative zones (cities, provinces, etc.). Of course, once a subscriber has healed from the virus, he/she would not "appear" on these maps anymore.
Standalone or backbone of a vertical solution, GeoHealth is definitely the solution all telecommunication operators should get to contain Covid-19.
Based on Intersec Technology, French Tier One SFR today partners with Public Hospitals of Paris and the French National Institute of Research in Science and Digital Technologies, making population movements flows data available for them. They were able to supply several datasets allowing hospitals to measure and visualize population movements between Paris region and the other French regions during March. These insights allow a better understanding of population movements’ mechanisms together with establishing correlations with epidemiological risks on the French national territory. The aim for both AP-HP and INRIA is to refine their awaited charge prediction models for both Parisian hospitals and Regional hospitals. And it is just one possible application of the solution!