What is this?
Viral Trails is a mobile app designed to help reduce the spread of COVID-19 by informing users if their location crossed paths with the an infected individual who has shared their historical data with us. This will help us selectively quarantine and seek testing and treatment of anyone exposed much earlier to significantly reduce the spread of infection and human health effects, reducing negative economic impact, and returning us back to normal faster.
How does this work?
Even while we are social distancing, we may need to go outside for food, necessary work, or other tasks. There we will inevitably encounter other people. And while we will do all the right things like wash our hands, keep our distance, etc. we may still cross paths an infected individual who is not yet showing symptoms but may still be contagious.
When that person is a user of Viral Trails and shows symptoms or tests positive, they will consent to upload their anonymous location info from the last two weeks, and anyone else using Viral Trails can compare their location log automatically (without it leaving their device) to see if they were in the same place at the same time as anyone who has since tested positive. If you were, you may self quarantine or opt to get tested to prevent further spread - freezing the virus in its tracks.
Doesn't everyone need to have this app for it to work?
The more people that have the app the better (so share this as much as possible) but... even if only 1% of people in your area use this, that will help us prevent those 1% of cases from spreading exponentially. If each infection spreads to 1% fewer people, over weeks, that compounds to thousands of prevented cases. By tracking your location and reporting it if you are sick to health authorities like the CDC, the data collected by this app will not only help other users of the app, but will aid in traditional contact tracking measures.
Has this been done before?
Yes. South Korea, Israel, and other countries have used location data to track the spread of the virus. In this case, we are taking the measures a step further to help ensure privacy of your location data which doesn't leave your device unless you consent to share it with our infection registry if and only if you test positive for the virus or if you explicitly choose to share it for aggregated research of the impacts of the virus and where health care resources should be deployed.
Does this work 100%?
If you cross paths with someone with the virus, Viral Trails will be your first line of information. You will be informed when and where you crossed paths with the virus to make a further risk assessment. This info does not mean you are infected, but will help you self-isolate and prevent further potential spread until you can confirm you are healthy. On the flip side, Viral Trails may not always see the full picture - you may catch the virus from a surface touched by someone who was sick hours before so it's still important to practice good hygiene and caution.
Does this work with other contact tracing solutions?
Yes. Unlike Bluetooth contact tracing which requires both parties to have compatible apps to "see" each other, location data has real world significance which can be used even within a traditional contact tracing setting. By looking at where an infected individual has been, authorities can determine who else might have been there even if they don't have Viral Trails installed. It also means authorities can go back and upload locations where infected users who didn't have the app were and alert users of Viral Trails in an efficient way. Viral Trails will also implement the Apple/Google Bluetooth Contact Tracing API when it becomes available to work with the larger ecosystem.
What security measures are in place?
Some solutions have been criticized for being able to "game" self-reporting and cause panic. Location data has a very nice property in that it is hard to generate real looking trips without actually taking the time and effort to travel them. That means one bad actor can only submit a full set of bad data once every 2 weeks per physical device. API protections and basic analysis of submitted pathways prevents most systematic abuse. Lastly, data shared by the API is protected by rotating keys which prevent it from being scraped and abused. A later implementation may use fully homomorphic encryption to enable secure multi-party computation without revealing any real data.
Want to get involved, or contact us?
Fill out this contact form.