sunemai

Your comments are very interesting, and I'd like more input/thoughts about such issues.

However, from your description of the Alice/Bob part, I'm still confused whether you realize that nobody is (or can be) quarantined in the proposed scheme?

The case you describe would, as I understand it be: "Bob is then advised by his own app not to visit his parents and be careful about not infecting others". A bad joke, sure, but not something that can "send him into quarantine", since nobody on the server side knows that he is in any danger.

I may be missing your point here - but if I am, please clarify and bare with me;-)

The privacy issue is complex. My take on it is quite simple:

I personally wouldn't mind if my (possibly randomized) BT address was made public, potentially identifying me as currently COVID-19 infected - heck, we will probably all have been on that list in a year or two as things stand today.

What someone (not me, of course) would probably mind, however, is if the geolocation data was used instead - and it showed what clubs I went to during the last 14 days.

The "buy in" issue is why I consider using Denmark (which is where I live) as an initial test area before internationalization: Danes are generally well educated, and they have a high trust in public authorities compared to many other countries. The use of open source software is also widespread in Denmark.

So an app/server solution developed with 100% transparency on GitHub could easily be pushed to everybody by public authorities - under the critical assumption that we can both make it work technically, and make it completely anonymous.

Just for reference, the same "scale problem" is also the challenge with ordinary vaccines, as I understand it:

They don't really need to give 100% protection to each individual. The critical point is that if a large enough percentage of the population has the vaccine, the R_0 parameter of the given disease will drop below 1, and the exponential spreading turns into an exponential decline.

The point is that, currently, people are "tracked down" by medical authorities and put in quarantine.

The AVICOT-APP concept is opposite: People would (should) like to know if they represents a danger to, say, their elderly parents, who may die if infected. Thus, there is no need for "the system" to be able to track or identify them at all. Only you, the end user, will ever know if you are in high risk.

I've heard that everybody is connected with a surprisingly low number of intermediate steps. Maybe now would be a good time to try that out;-) So, if you know anybody...

If you are able to check whether you've been in a "dangerous" space-time box yourself, then you are probably also able to check where these boxes are, I suppose? One way would be brute force hashing all possible space-time blocks and comparing with the public hashed list. This mechanism was why I abandoned the space-time box approach to begin with - but if you have an idea for a workaround, please get in touch!

Hi, I'm the developer behind https://github.com/AVICOT-APP/Documentation

Thank you all for your valuable input - and sorry for the late reply: I posted under my own account (https://news.ycombinator.com/item?id=22462696) - but this thread initiated by a friend was the one to be picked up.

I'll try to address some of the issues brought up under each comment later, but here is a few things to consider first:

The MAC address randomization is a key issue here. For the "level 0 list", though, it might not be that much of a problem: If the app also keeps track of its own actual MAC address(es), then these are what would be published in the "level 0 list", along with the corresponding time-index. I'll update the GitHub description to take this into account, but please help me answer some of these questions first (or, better, write me an email, so you can contribute directly to the AVICOT/Documentation repo) :

Does the randomization happen at each new invocation of Bluetooth?

Is it coupled with the "location services" as suggested by jackweirdy below?

Another important point is that of using location tracking instead, as suggested by c22. I had this exact scheme as the initial AVICOT concept, but abandoned it because even with one-way hashing of the coordinates of the space-time boxes, the locations (and thus identity) of the infected person can easily be calculated if the (hashed) "dangerous" space-time boxes are publicly available.

One really good point of using space-time location boxes is, that you can assign an infection risk to being at the same location, say 3 hours later (e.g. use an exponential risk decay with some suitable half-life parameter). This is not possible with the Bluetooth IDs. But unless we solve the anonymity issue in some way, I think the Bluetooth way is easier.