Overview for Participants: The Safe Blues Experiment at UoA
The Safe Blues campus experiment aims to evaluate the predicative power of the Safe Blues system. All participation is on a voluntary basis, see the participant information sheet. Participants are mostly students, however the experiment is also open to staff, or regular visitors of the University of Auckland City Campus. Participants run the Safe Blues app on their Android phone as they go about their normal day to day activities on campus. A participant's app then communicates with the apps of other participants via Bluetooth to create a virtual safe virus spread simulation while recording anonymized, aggregate data according to the Safe Blues protocol.
The app does not require any participant input, participant identity, and does not scrape any information from the participant's phone. The app does not track the location of participants with the exception of indicating if the participant is within the campus area or not. All information recorded is aggregate information indicating how many virtual safe viruses have 'infected' a participant's phone. This information is never cross-validated with private participant information and hence the experiment is a fully anonymized experiment. The only private information stored for each participant is the collective usage hours of the app on campus and this is solely for purposes of the prize draws.
As participants run the Safe Blues app, their app transmits and receives Bluetooth signals that mimic the spread of an epidemic. There are many types of 'strands' of such virtual safe epidemics, and the live simulation of all of the epidemics happens in parallel. The app also reports the collective number of 'virtual infections' to a server - however this data collection is aggregated and anonymized in such a way that the virtual safe virus strands on any specific phone are not recorded.
This is the first ever test of Safe Blues. The hypothesized benefit of such a system to empower public health authorities to better control, measure, and mitigate epidemics such as COVID-19. See also this short video (with much more information on the main Safe Blues website). Our aim in the experiment is to better understand how such a system would work in a real life scenario.
The Safe Blues experiment does not involve any biological (real) viruses. While the Safe Blues system is ultimately designed to estimate the evolving rate of transmission of biological viruses using the virtual safe viruses as a proxy, this is not the case for the experiment.
An additional output of this experiment will be an anonymized open sourced dataset, available to researchers and machine learning enthusiast. The data set will include time stamped aggregated counts of multiple Safe Blues strands, recorded during the experiment.