On March 20-22, 2020, the German federal government sponsored a hackathon called #WirVsVirus (#WeVsVirus) that centered on finding solutions to social challenges related to the Covid-19 crisis. There was an enthusiastic response and over 42,000 people registered to take part. I was part of a small team from Stanford that had gathered (virtually) to work on this hackathon. We chose to work on the challenge of mass emergency notification systems for citizens living abroad. This article describes my experience with the hackathon, our team’s work and our outcomes, and our next steps. It also provides an overview of the organizational structure of the hackathon from the perspective of a participant.
Challenges in Planning a Hackathon with 42,000 Participants
There have been many hackathons centered around a particular cause, as well as hackathons sponsored by local and federal governments, but this must be considered one of the largest hackathons to date. Coordinating such a massive event was, in and of itself, a significant accomplishment. In the days leading up to the hackathon, there was an open call for people to submit their ideas for challenges for the hackathon; in the end, there were over 3,000 suggestions. The hackathon organizers edited that list down to approximately 1,000 challenges, which were categorized into 48 categories and posted in an Airtable, which is a type of database-spreadsheet hybrid. The plan was for the hackathon participants to sign up as individuals and then organize themselves into teams based on the challenge that they chose to work on. However, since the organizers had planned for a few hundred participants, and in the end the hackathon had over 42,000 registered participants, there were a few technical issues.
Perhaps the main lesson for others planning to hold a massive online hackathon is to build as many backup communication systems as possible into their planning. The hackathon organizers used e-mail as their primary communication channel to inform participants about the hackathon schedule and to distribute links to various sites such as Airtable, DevPost, Slack, YouTube, etc. that were used for other functions. The e-mail notifications worked well, but Slack, which was the main channel for communication among teams and individual participants, was overwhelmed by the number of participants. The organizers managed to pull it together after a delay of several hours; the best workaround turned out to be adding all 42,000 participants to the Slack channel manually (which was only made possible by the numerous volunteers on the organizing team). During the delay, the most reliable source of information and updates was through the official hackathon twitter channel.
The hackathon schedule was clear and well-planned. The opening ceremony on Friday, and the closing ceremony/dance party on Sunday, were broadcast live on YouTube. The final submission for the hackathon was due at 6:00PMGerman time/9:00AM California time on Sunday. The submissions on DevPost were also due at that time. All teams had to upload a 2-minute video to YouTube describing their solutions. Submissions were evaluated on the following criteria: Gesellschaftlicher Mehrwert (social added value); Innovationskraft (innovation); Skalierbarkeit (scalability); Fortschritt während des Hackathon (progress made during the hackathon); and Verständlichkeit der Lösung (the solution’s understandability). The hackathon was held in German, which automatically limited the participants to those who knew German, although there were some people who offered to translate for non-German speakers.
The Stanford Team and Our Project
The leader and coordinator of our team was Ruth Elisabeth Appel, a Ph.D. Student in Media Psychology in the Stanford University Department of Communication. Other team members were current Stanford students, staff, alumni, or those had worked at Stanford in the past and still maintained their connection. In addition to the technical issues caused by so many participants, our team faced the hurdle of the time difference. When we began gathering to work at the official hackathon start time of 9:00AM California time/6:00PM German time, we discovered that the launch had been delayed. Eventually, after a rousing YouTube launch, we set to work. Our team narrowed the list of challenges in Airtable down to our top seven choices, which we discussed in a face-to-face Zoom call before choosing challenge #313 (below, in my translation). We chose this challenge because most of the Stanford team were German or Austrian citizens who themselves had recent experience with the ELEFAND system and expressed the desire to improve on it.
Challenge #313: Mobile information platform for Germans living abroad
Problem: Efficient communication of information to Germans who work abroad and cannot return to Germany owing to their positions. Almost 1.9 million Germans work for German companies or institutions abroad […] The following information is particularly important and is difficult to obtain in the country of residence owing to language, etc.:
- Emergency contacts at the embassy
- Procedures in the event of suspected or confirmed cases of COVID-19
- Hospitals and outpatient clinics that can test for and treat COVID-19 if necessary (with adequate health standards)
- Risk assessment for individual countries from the German Federal Foreign Office
- Evacuation plans and information on local travel regulations
- Local regulations for quarantine measures
- Prevention measures and procedures for quarantine (especially for families)
- Information on health insurance, etc., such as when insurances offices in Germany are closed
- Current information on developments in each country
- Embassies / Consulate should be able to use a direct communication channel in an emergency (such as an evacuation situation, as in Wuhan).
At the present moment, embassies and consulates are overwhelmed by a particularly large number of inquiries and cannot adequately respond. In addition, there is great uncertainty among German expats abroad, as the information situation in some countries is very non-transparent.
Challenge: How can comprehensive information and updates be efficiently transmitted to Germans abroad in order to improve the security situation of Germans abroad and to relieve consulates? ELEFAND (the current platform) doesn’t work very well and there are more efficient alternatives (e.g. WhatsApp).
As part of the registration process, we chose a team and created a project page on DevPost. Here is our team description (translated):
Group name: ELEFAND 19.0 (allusion to the current platform for Germans abroad, ELEFAND, and 19 as in COVID-19)
Group members: Ruth Appel, Andreas Zoellner, Juliana Abramovich, Sebastian Schneider, Daniel Gruen, Kathleen Smith, and Jan Sokol
Group description: Our goal is to enable better communication between German citizens living abroad and German governmental agencies–particularly in emergency situations. In recent days, we have had direct experience with the crisis communication through channels such as ELEFAND, which is sorely in need of improved usability, content, and information distribution. We would like to improve/replace ELEFAND with more efficient/more accessible alternatives (such as through better integrated information channels and/or new channels such as an app).
One characteristic of all hackathons is their concise time frame; the goal is to produce within an intensively-focused window of time a prototype that shows that the concept can work, with the functioning at scale and real-world implementation following later. Our team began with a broad challenge and we also had to keep in mind that our goal was to produce a result by the end of the hackathon. Our brainstorming session and internal documents reflect the scope of our research activities as we assessed existing official and unofficial crisis communication channels in Germany and other countries, explored the stakeholders, created use cases for the communication system, evaluated integration with other existing platforms, and defined the team goal—all over the span of two days.
The importance and immediacy of our challenge was immediately confirmed: on the second day of the hackathon (March 21), there was a news article in the German magazine ComputerWoche about how the current ELEFAND system had crashed under high demand. In response, the German software company SAP was springing into action to build an app that could be used to coordinate evacuation efforts for all the German citizens stranded around the world.
That same day, Ruth Appel was able to arrange for us to speak with Marietje Schaake, former EU parliament member, and current fellow at the HAI at Stanford, about this challenge and our approach to it. She encouraged us to think about what features our proposed solution should have, regardless of the form it took.
The final submission for the hackathon was due at 6:00PM German time/9:00AM California time on Sunday, March 22. The submissions on DevPost were due at that time. Our video (in German) is posted to YouTube, and here is our DevPost submission.
Our team member Andreas Zoellner built a prototype (andreaszoellner.de) to demonstrate our features; we also began a spreadsheet comparing the various systems in use by countries around the world.
The members of my team and I are interested in continuing to develop some of the work we started during those two fast-paced days. It was very helpful to feel like we were still in contact with other humans, that we were all “in this together,” and to put our backgrounds and expertise to the test in a real-world situation.
One of our main recommendations was to include a built-in feedback loop into whatever emergency notification system is in use, since for the rest of the members of my team who were all German/Austrian citizens, it was difficult to know if ELEFAND was still working since they never received any notifications and there was no way to contact the administrators of ELEFAND, leave feedback about the system, or even to reach out to other users.
As part of our research, we started a spreadsheet surveying how governments around the world have addressed the challenge of communicating with their citizens outside the country (traveling, working abroad, etc.), with the goal of comparing these systems and their effectiveness. This comparison provided useful input but we were not able to fully complete it within the time frame of the hackathon. We would like to continue expanding upon it—if you would like to contribute, please contact me for more details.
If you are curious to learn more about the solutions proposed by other teams, you can access the initial Airtable, the Slack channels, the DevPost pages for all projects, and the YouTube channel via the English translation of the #WirVsVirus page.
Stay safe, everyone!
Kathleen Smith, PhD, MLS
Curator, Germanic Collections
Stanford University Libraries