NZ COVID counts increase, yet a digital experiment at Auckland Uni demonstrates lockdown works
Delta cases were discovered in Auckland New Zealand earlier this week and the country quickly moved into lockdown. The first case was discovered on Tuesday Aug 17, and now on Saturday Aug 21 there are already over 20 confirmed cases. Since the initial months of the COVID-19 outbreak, New Zealand was considered a COVID safe haven. Rapid initial responses crushed the pandemic to date and there have been less than 3,000 total cases over all of time. For a population of just about 5 Million people this is just about the lowest infection rate in the world.
However the recent delta outbreak is very concerning. This is especially the case given the exponentially rising numbers in parts of nearby Australia.
Meanwhile, there is an ongoing experiment at the University of Auckland. The experiment uses a framework called Safe Blues, where virus-like tokens are spread via Bluetooth between cellular phones of participants. In this ethics approved experiment, students run the Safe Blues app in the background of their Android device. When students are physically nearby, their phones exchange tokens and this mimics the spread of epidemics.
The goal Safe Blues is to gain better information about human contact trends while preserving privacy. This is not for tracking individuals but rather for assessing how virus spread may respond to social distancing measures. Ultimately statistical techniques and AI can use real-time Safe Blues measurements to predict the state of the virus. However, now in view of the flash Auckland lockdown, the Safe Blues experiment is already giving very promising signals.
The experiment involves hundreds of digital token virus variants with varying (simulated) disease attributes. For example, one set of such (safe) viruses has a random incubation period with a mean of 3 days, and a random disease duration (during which a phone can infect other phones) with a mean of 10 days. This is somewhat similar to COVID-19. The above figure illustrates the trajectories of 12 repetitions of such a virus.
A random subset of the population was infected with this virus three weeks ago, and at the day of the New Zealand lockdown an average of about 25 participants were infected with each such repetition. This is not so different from the estimated state of the delta strain on that day.
Now, 4 days into the lockdown, the system’s live measurements are already showing that the number of exposed participants (incubations) is decreasing. However, the number of infected participants still isn’t.
With real COVID cases, one cannot measure exposed cases, and hence the effect of the lockdown is still not visible on the actual case numbers in New Zealand. However, Safe Blues allows us to immediately see the actual affect, albeit on simulated viruses in people’s phones. In the coming days we’ll continue to track the state of the Safe Blues virtual safe epidemics in parallel to actual case numbers reported from New Zealand.