There has been a lot of talks about contact tracing to help reduce the spread of COVID-19 virus. But the process seems to be very labor intensive and relies heavily on individuals’ recollections. Apple & Google joined force and developed a solution, but from what I’ve learned, it was based on formal “contacts” and may still requires a lot of manual work.
As I think the problem at hand, I wondered if there is a simpler way to do contact tracing anonymously without much manual tracing work, and if there are existing tools that can be used.
I think I’ve found a simple, elegant (and cheap) solution: Google Analytics.
Here is the concept:
Google Analytics already track user anonymously. We just need to generate a unique user ID to identify each user.
We record GPS locations every 5 minutes from the smartphone, tag the unique user ID and submit/store the information in Google Analytics.
The raw Google Analytics events data is accessible via Google BigQuery.
When a person was diagnosed with COVID-19, he or she can submit the unique ID from the smartphone to a “tracer bot”
The bot will then do a query in BigQuery to find all the “contacts” based on GPS and timestamp of the events.
Contacts with high risks can then be notified anonymously via push notifications.
Here are the Google Analytics accounts I’ve setup for this concept:
GA Tracking ID: UA-164925046-1