Why is (big) phone data so valuable in combatting the COVID-19 pandemic?
Since the start of the epidemic, data has been used to help governments limit the spread of the virus. In Europe, the European Commissioner for Internal Market and Services Thierry Breton has, along with the scientific research community, been calling on telecoms operators to join the effort. Data is an extremely valuable tool for anticipating the spread of the virus. In France, Orange acted very early on to support medical research while fully respecting personal data privacy.
What is data used for?
Data is only valuable if it can be compared, associated or aggregated with other data. The major value of telecoms data, and in particular population location data, is to establish accurate statistics about people’s movements, since people tend to take their mobiles with them when they go anywhere. What percentage of the population left large cities just before lockdown measures were introduced? Where did this population then move to across the country? This information is valuable for scientific knowledge, which is why our researchers are working closely with the teams at Inserm on mathematical models to anticipate how the epidemic is evolving, using the most accurate and “real time” statistical and demographic data possible.
These statistics are also of interest to the French authorities, which are deploying medical resources throughout the country. By informing them that 17% or 1.2 million Parisians left the Île-de-France region in the days preceding and following the start of the lockdown, while the Yonne region gained 10%, it can help them better calibrate the hospital system or anticipate the required medical equipment deliveries. We are also able to confirm whether social distancing measures are being respected at a population and individual level so they can adjust the strategy if needed.
How does it work?
Our objective is to establish useful statistical tools to help combat COVID-19. To do so, we’re taking advantage of our Flux vision solution, which was launched after a 10-year R&D programme and developed in compliance with the criteria established by CNIL.
Information about people’s mobile phone connections, whether customers or other users on our network such as tourists, can be collected and aggregated in real time from relay antennas. The data is not stored and we’re only keeping the overall total, so there is no risk that it will be used at a personal level.
The initial results are then adjusted using conventional statistical tools to allow projections to be made. This means that from a sample of 27 million devices connected to our network, we’re able to extrapolate the population’s behaviour as a whole and establish their movements from one region to another. We then pass this data onto public authorities such as Inserm, prefectures and health authorities.
What about individual applications in the future?
Some countries have chosen to use mobile apps and user data without asking for people’s consent. In Taiwan for example, authorities are alerted if a quarantined individual leaves their home. Elsewhere, this type of practice needs to be carefully considered in regards to General Data Protection Regulation (GDPR) compliance.
Other governments, on the other hand, have launched volunteer-based individual protective apps. This is the case in Singapore and South Korea in particular. The app detects via Bluetooth how close the user is to other people using the app. A user who tests positive to the virus can indicate it via the app. This information is then sent to everyone they have been in contact with so they know they have been exposed to a risk of infection. This will be a particularly useful tool towards the end of the infection period to prevent the epidemic from restarting. This is the type of app we are considering. We’ll be paying full attention to recommendations issued by CNIL and European authorities. In our opinion, such an app must meet the following three conditions for mass adoption by users, which will be the key to its effectiveness:
- comply with personal data regulations following discussions with CNIL experts;
- be deactivated and all stored contact data deleted at the end of the crisis;
- be made available in open source so that the developer community can verify its security and integrity.
The new Analysis, Research and Expertise Committee Care(Comité analyse, recherche et expertise), which aims to advise the French government on ways to best combat the epidemic, will also be deciding on the digital strategy to deploy.