The virus will be with us for several months. That's why this question has been on our minds most as a team: How do we live with the corona virus as comfortably and safely as possible during this time?
**Why is it so important as a virus carrier to avoid contact with healthy people? The COVID-19 virus is highly contagious. While I infect about 1.4 more people with an influenza infection in autumn, I pass on the COVID-19 virus to about 3 people. These numbers seem similarly low at first. However, each person infected by me passes on the virus with the same probability. After only 10 transmissions from person to person, I was the first influenza carrier to infect 29 other people, but with COVID-19 I have already infected 59,049 people.
PANDOA in everyday life
I would like to be able to move as freely as possible and still be informed in good time about risk contacts and thus an increased risk of infection. PANDOA is the interface to determine my own risk on the one hand, and to inform the health system about new sources of infection as quickly as possible on the other. We attach great importance to an easy-to-use tool that allows me to keep control over my data at the same time.
**How do I know my own risk? I will be informed by PANDOA if I have had contact with a risk person in the past few days. In return, I will make my whereabouts available in anonymous form if I fall ill myself.
**How is the health care system prepared for new outbreaks of infection? By making the whereabouts of persons at risk anonymously available, PANDOA can quickly identify sources of infection and enable doctors and hospitals to respond in a regionally specific manner.
The team behind PANDOA
Two days ago, a colourful team of more than 20 developers, project managers, doctors and graphic designers met for the first time in virtual space and 48 hours later we can present you a runnable app with open interfaces.
How does PANDOA work?
The app is unobtrusive and intuitive to use. PANDOA continuously records my motion data and stores it locally on my smartphone for 14 days*. In addition, the most clinically critical parameters of my health status are queried daily: body temperature, quarantine status, infection status. Together these data are stored locally on my smartphone without being shared. According to my voluntary health status information, the app gives country-specific recommendations for further action. If I fall ill myself, I can share my whereabouts of the last days with all other users in anonymous form. This is important, because I could have spread the virus before the signs of illness appeared. So I can sensitize my fellow men for the appearance of disease symptoms.
This is important:
- The recording of the data on my smartphone is temporary.
- I decide myself when I want to share which data in anonymised form.
- It is not possible to track individual persons (in the video, this illustration serves to show the actual functionality of the algorithm). However, I can identify regions with high infection density on a map. *Period of storage corresponds to current knowledge about the maximum incubation time of COVID-19.
The open-source interface PANDOA
The app is open-source and can be supplemented with information from other applications. Already developed are the interfaces to precision of my risk by connecting self-diagnosis tools (e.g. CovApp of the Charité), as well as the support of models on the future spread probability of the virus (e.g. SIR models) and the possibility to voluntarily provide my data for epidemiological research (e.g. machine learning analysis of pooled data sets). For this we are already in contact with the developers of some tools, also from the #WirVsVirus challenge (see below).
Technical details about PANDOA
Multiplatform map for iOS and Android
- GPS based tracking and with local storage
- (daily) health status query with body temperature, quarantine and test status (positive/negative)
- Consultation with the users:in order to share the data, should the own health status deteriorate
- Download of movement data of infected persons and reconciliation of movements
- Warning if your own transaction data overlaps with that of infected persons
We have currently opted for the geotagging method for detecting overlapping contacts. This is the only way to generate data on regional distribution and for epidemiological questions. Furthermore, we do not see the continuous contact matching across operating systems as guaranteed with alternative methods (e.g. Bluetooth), especially if User:in Bluetooth is not explicitly enabled.
- MongoDB for storage and output of transaction data and health status
- Cloud architecture AWS with S3 storage
- MongoDB scalability already implemented