Smartphone contact tracing has failed everywhere

> It would appear that no government has thought this out properly yet.

This seems to be true at least in India, where the app is a centralized one that doesn’t use the Apple/Google framework The app tracks the location of users at all times (it asks for that permission) and hence cannot use the APIs since Apple and Google prohibit such apps from collecting location information. Since it requires background location access and Internet connectivity, it’s a good addition to drain batteries quicker.

The government also famously declared that the app would be open sourced so that it could be improved and issues fixed. That turned out to be a farce with the source released only for the Android version, and that too for an older version of the app with some code removed (the code on the Play Store was newer). They stopped releasing updates of the source. They didn’t release the source of the iOS app, which was supposed to be “shortly” (that was around two months ago). They didn’t fix the issues that people raised on GitHub. Yet the government claimed that the app was effective and put out some numbers on it being useful (without any backing data or any way to verify those claims).

It gets even worse. The app being a platform for contact tracing is only one facet. The goal that wasn’t widely disclosed was to have private companies build on top of the app and its server side data to create platforms on health data collected from hundreds of millions of people. Note that India doesn’t have a data privacy or data protection law, leave alone anything to control health related data in the hands of crony capitalists and opportunists.

Now all talk about contact tracing and its effectiveness is muted or nowhere to be seen, though the app is being pushed as a requirement to give people permission to travel and for other purposes.

The platform has become, and is continuing to become, what some people feared...and that’s getting farther from handling COVID-19 or a public health crisis.

Between them, they have exact positions of a vast majority of individuals, at least in western societies. They could have done useful contact tracing at the press of a button, even historically, for every major event such as large concerts, churches, night clubs which we know were the most important situations to contact trace. They could have done this with the data they already have.

This would however set a bad precedent on how this data is shared. Law enforcement is one thing, medial tracing is another. So instead they set out to build this BLE tracing toolkit, which is something that at best is intended for future use, when everyone has the requisite hardware and software installed. It wouldn't reach more than a minority of users today which severely limits its usefulness, all for the possibility of reaching into situations such as bus rides, where nobody really knows how contagious these situations are because we simply do not have enough knowledge of the virus yet. Real world measurements are also likely to be noisy (where someone carries their phone in their hand and another in their backpack with a water bottle) even if we could set a hard limit on a safe distance, which we can't.

So maybe we will get something useful in the future, for the price of not exposing the data already available today.

It is amusing that the BLE tracing toolkits were sold as a way of not having to expose your location data to Google and Apple, the exact entities who already have fine grained location data on most people.

Its only recommended for more serious viruses like Ebola.