The problem WeCare solves
As the outbreak of COVID-19 continues to spread throughout the entire world, more stringent containment measures from social distancing to city closure are being put into place, greatly stressing people we care about. To address the outbreak, there have been many ad hoc solutions for symptom tracking (e.g., UK app), contact tracing (e.g., PPEP-PT), and environmental risk dashboards (covidmap). However, these fragmented solutions may lead to false risk communication to citizens, while violating the privacy, adding extra layers of pressure to authorities and public health, and are not effective to follow the conditions of our cared ones. Until now, there is no privacy-preserving platform in the world to 1) let us follow the health conditions of our cared ones, 2) use a statistically rigorous live hotspots mapping to visualize current potential risks around localities based on available and important factors (environment, contacts, and symptoms) so the community can stay safer while resuming their normal life, and 3) collect accurate information for policymakers to better plan their limited resources. Such a unified solution would help many families who are not able to see each other due to self-quarantine and enable early detection and risk evaluation, which may save many lives, especially for vulnerable groups. These urgent needs would remain for many months given that the quarantine conditions may be in place for the upcoming months, as the outbreak is not reported to occur yet in Africa, the potential arrival of second and third waves, and COVID-19 potential reappearance next year at a smaller scale (like seasonal flu). There is still uncertain information about immunity after being infected and recovered from COVID-19. Therefore, it is of paramount importance to address them using an easy-to-use and privacy-preserving solution that helps individuals, governments, and public health authorities. The closest solution is COVID Aggregated Risk Evaluation project, which tries to aggregate environment, contacts, and symptoms into a single risk factor. WeCare takes a different approach and a) visualizes those factors (instead of combining them into a single risk value) for more tangible risk communication and b) incentivizes individuals to regularly check their symptoms and share it with their Care Circle or health authorities.
WeCare is a digital platform, both app and website. Both platforms can be used separately, and with freedom of choice towards the user. The app, however, will give users more information and mobile resources throughout the day. Our cross-platform app enables symptom tracking, contact tracing, and environmental risk evaluation (using official data from public health authorities). Individuals can add their family members and friends to a Care Circle and track their health status and get personalized daily updates. In particular, individuals can opt-in to fill a simple questionnaire, supervised by our epidemiologist team member, about their symptoms, comorbidities, and demographic information. The app then tracks their location and informs them of potential hotspots for them and for vulnerable populations over a live map, built using opt-in reports of individuals. This map is accessible on the app and our website. Moreover, symptoms of individuals will be tracked frequently to enable sending a notification to the Care Circle and health authorities once the conditions get more severe. We have also designed a citizen point, where individuals get badges based on their contributions to solving pandemic by daily checkup, staying healthy, avoiding highly risky zones, protecting vulnerable groups, and sharing their anonymous data.
Our contact tracing module follows guidelines of Decentralized Pan-European Privacy-Preserving Proximity Tracing (PEPP-PT), which is an international collaboration of top European universities and research institutes to ensure safety and privacy of individuals.
What we have done during the weekend
We have updated the app-design this week (w.21). New contacts with Chile and Singapore. We have also made some translation work with the app. Shared more on social media about the project and also connected to more people on slack and LinkedIn.
We have consolidated the idea and validated it with a customer survey. We then developed a new interface for website and changed the python backend to make it compatible with the WeCare app. We have also designed the app prototype and all main functionalities:
- Environment: We have developed the notion of hotspots where we have developed a machine learning model that maps the certified number of infected people in a city and the spatial distribution of city population to the approximate number of infected in the neighbourhood of everyone.
- Contact tracing: We have developed and successfully tested a privacy-preserving decentralized contact tracing module following the (PEPP-PT), guidelines.
- Symptoms tracking: We have developed a symptom tracking module for the app and website.
- Care Circle: We have designed and implemented Care Circle where individuals can add significant ones to their circle using an anonymous ID and track their health status and the risk map around their location. You can change what info you want to share with Care Circle during the crisis.
The app is very easy-to-use with minimal input (less than a minute per day) from the user. We are proud of the achievements of our team, given the very limited time and all the challenges.
Challenges we ran into
EUvsVirus Hackathon Challenge opened its online borders recently to the global audiences which brought together plenty of people of different expertise and skills. There were challenges that we faced that were very unique, as we faced a variety of communication platforms on top of open-source development tools.
Online Slack workspaces and Zoom meetings and webinars presented challenges in forms of inactive team members, cross-communications, and information bombardment in several separate threads and channels in Slack and online meetings of strangers that are coordinated across different time zones. In developing the website and app for user input data, our next challenge was in preserving the privacy of user information.
In the development of a live map indicating hotspot regions of the COVID-19 real-time dataset, our biggest challenge here was to ensure we do not misrepresent risk and prediction into our live mapping models. We approached Skill Mentor Alise. E, a specialist in epidemiology, who then explained in greater detail that the proper prediction and risk modelling should take into account a large number of factors such as population, epidemiology, and mitigations, etc., and take caution on the information we are presenting to the public. Coupled with the lack of official datasets available for specific municipalities for regions, we based geocoding data mining of user input by area codes cross-compared with available Sweden cities number of fatalities, infected and in intensive care due to COVID-19.
The solution’s impact on the crisis
We believe that WeCare would help many families who can see each other due to self-quarantine and enable early detection and risk evaluation, which may save many lives, especially for vulnerable groups. The ability to check up on their Care Circle and the hotspots around them substantially reduces the stress level and enables a much more effective and safer re-opening of the communities. Also, individuals can have a better understanding of the COVID-19 situation in their local neighbourhood, which is of paramount importance but not available today.
The live hotspot map enables many people of at-risk groups to have their daily walk and exercise, which are essential to improve their immunity system, yet sadly almost impossible today in many countries.
The concept of Care Circle motivates many people to invite a few others to monitor their symptoms on a daily basis (incentivized also through badges and notifications) and take more effective prevention practices. Thereby, WeCare enables everyone to make important contributions toward addressing the crisis. Moreover, data sharing would enable a better visual mapping model for public assessment, but also better data collection for the public health authorities and policymakers to make more informed decisions.
The necessities to continue the project
We plan to continue the project and fully develop the app. However, to realize the vision of WeCare we need the followings:
- Social acceptance: though being confirmed using a small customer survey, we need more people to use the WeCare app and share their data, to build a better live risk map. We would also appreciate more fine-grained data from the health authorities, including the number of infected cases in small city zones and municipalities.
- Public support: a partnership with authorities and potentially being as a part of government services, though not being necessary, to make it more legitimate. This would increase the level of reporting and therefore having a better overview and control of the crisis.
- Resources: So far, we are voluntarily (and happily) paying for the costs of the servers. Given that all the services of the app and website would be free, we may need some support to run the services in the long-run.
The value of your solution(s) after the crisis
The quarantine conditions and strict isolation policies may still be in place for upcoming months and year, as the outbreak is not reported to occur yet in Africa, the potential arrival of second and third waves, and possible COVID-19 reappearance next year at a smaller scale (like seasonal flu). Therefore, we believe that WeCare is a sustainable solution and remains very valuable after the current COVID-19 crisis.
The URL to the prototype
We believe in open science and open-source developments. You can find all the codes and documentation (so far) at our
https://www.facebook.com/wecareteamswedenhttps://www.instagram.com/wecare_teamhttps://www.linkedin.com/company/42699280https://youtu.be/_4wAGCkwInw (new app demo 2020-05) Interview: https://www.ingenjoren.se/2020/04/29/de-jobbar-pa-fritiden-med-en-svensk-smittspridnings-app
Try It out
node.js, python, react, vuejs