Give it a try
How can we use validated, real-time data to detect and predict virus outbreaks of infection early to increase transparency for government and citizens as well as to improve the management of this infectious disease?
What it does
Visualize and see the current corona-outbreak situation in our area, region or country and simulate its spread in the nearest future according to parameters like “Ausgangssperre”. This shows how the situation is likely to change in the future and what happens if people follow the stated rules.
How we built it
We have used a state-of-the-art technology stack.
- Backend: Python, Flask
- Database: MongoDB
- Datasets: Corona cases and deaths (provided by Robert-Koch-Institut), German city statistics (provided by destatis.de), German city geolocation data (provided by OpenDataSoft)
- Hosting: AWS S3 (frontend) & EC2 (backend & database)
- Simulation: modified SEIR model (plain Python) & Data Associated Recurrent Neural Network (Python, PyTorch)
Challenges we ran into
- We are a very diverse and interdisciplinary group of individuals with diverse areas of expertise located primarily in the Karlsruhe, Munich, Hamburg area. The members when coming together did not know each other well. Hence, we had to go through a team forming process first to create an “agile” team.
- Since we did not know each other right from the beginning, we had a major challenge to find the perfect responsibilities for everyone. Thus, we enabled smooth group hopping by keeping all conversations public in the designated Slack threads within our slack channel.
- The team was quite large. Thus, we have defined a common vision with the team. Based on the vision, we have organized ourselves in four different subgroups based on different parts of the final platform and had regular touchdown meetings in the big group every other hour for consolidation and synchronization of the results. Also, we had a stream that focuses on general organization of touchdown meetings, communication and storytelling.
- From Saturday to Sunday our AWS-Database was “hacked” and deleted. We had to reset it and have established some security mechanisms.
Accomplishments that we're proud of
- We were quite successful bringing a large number of highly experienced individuals on diverse locations together in a well performing team
- We managed to create a very positive and constructive atmosphere while maintaining momentum on MVP-Development, organization and communication.
- We have solved a very complex mathematical problem during a very limited timeframe of 48h that provides value to both, government officials as well as citizens.
What we learned
- We have learned that despite the current corona crisis we as a team that did not know each other beforehand can create a solution that can create an impact and value for government and society. What would be possible and what could happen on a larger scale with a little funding and even more expertise?
- We are proud to see what can be achieved with an interdisciplinary, intrinsically motivated team without any funding in a 48h time slot
What's next for the project
- Incorporporate more data sources (e.g. world-wide data).
- Add and model additional factors that can limit the number of infections.