Coronicle - Gemeinsam tracken wir den Virus!
Inspiration
It's increasingly difficult to find out where an infected person has contracted the virus and who else she might have infected already. Right now this information is collected manually but with the current rise in numbers the officials are unable to keep track. The idea: Our smartphones are already enabled with precise location trackers so they can do this work for us! Coronicle ist a digital solution to identify infection chains and stop them in an automized, scalable fashion.
What it does
For the user, at first our app doesn't do much more than most fitness tracker: it records where you've been and when you've been there using your GPS data. That will be saved on your local device. There are two more ways to interact with our application. If you are infected, you can share your data to our database, where it will be stored in an anonymized way. If you want to check if you might have had contact to an infected person you can compare your tracked timeline with records from infected users, and find out if you you've had any overlaps.
How we built it
Our system relies on two systems. The first system is the smartphone application, which has been built using JavaScript. Data storage and transmission will be handled by SQLite. On the second system, the backend application is written in Java 13 on a Spring server. Both systems communicate via a REST api.
Challenges we ran into
A major challenge has been to define how a possible infection would be detected, e.g. finding reliable data on how persistent the virus is in a given environment. As SARS-CoV-2 is a relatively young virus there's little data on how the infection does spread, so we're now relying on general information (like distances) and survivability based on tests in lab environment.
Challenges we ran into
The main challenge was to come up with a design that respects the privacy of the user, while still providing the needed valuable information. We decided to only store the movement history of infected persons in the database. This data is anonymized and also dissociated, i.e. we only collect a mass of locations, without connecting the locations belonging to one infected user. We only need to know where the infected people have been collectively, and not where one individual infected user has been. When another user sends his location data to check whether she has been close to the virus, that data is not saved but only used for the calculation of the necessary information. This way, we believe to have found a good trade-off between privacy and utility.
Another major challenge has been to define how a possible infection would be detected, e.g. finding reliable data on how persistent the virus is in a given environment. As SARS-CoV-2 is a relatively young virus there's little data on how the infection does spread, so we're now relying on general information (like distances) and survivability based on tests in lab environment.
Accomplishments that we're proud of
Building a working prototype with front- and backend system in very little time and REMOTELY is a huge achievement for all participants. A huge thanks to all participants working on the project!
What we've learned
When it comes to tracking data and storing them it's very difficult handle such information without compromising personal information. We thought a lot about the main trade-off and realized that in a democratic society like Germany, stopping infection chains will require a lot of cooperation and solidarity from citizens. In emergency times, citizens should voluntarily share the data that is needed to solve the problem, rather than being forced or secretly spied upon (as is now happening in other countries).
What's next for Coronicle
We want to join forces with all the other teams that have been working on similar ideas. We need to take the best ideas from all these projects, unite them and work together on a solution that can solve the problem. Before we can roll out an alpha test we definitely have to find a GDPR-conform way to handle the tracked data of infected persons.
Log in or sign up for Devpost to join the conversation.