One of the 5 million Bluetooth tracing tokens being distributed to every adult in Singapore. Photo / Singapore Government
Debate over automated vs manual Covid tracking has reignited, with data showing a sharp tail-off in Kiwis using the NZ Covid Tracer app - just as Singapore distributes Bluetooth dongles to its population that will automate the process of tracking.
Diligent NZ Covid Tracer app use by one of the three people involved in the latest sub-cluster has been credited with speeding up contact tracing for their activity centred around Taupō and elsewhere.
But this week Dr Andrew Chen, a researcher at University of Auckland-based Koi Tū: The Centre for Informed Futures, highlighted that Ministry of Health data shows a sharp fall-off in QR code scans.
About 2.6 million people (around 55.4 per cent of those aged 15-plus) have downloaded the app - which is very good, by international standards. But after a steady increase in use after the Auckland outbreak in early August, it's tailed off again. Most of us are now not reaching for our phones to scan the QR poster before walking into our local supermarket.
NZ COVID Tracer daily scan counts are dropping fast. We didn't even have to drop alert levels before people have become complacent - behaviour change hasn't been sustained and people are returning to old behaviours. pic.twitter.com/hlKLoURzsA
— Andrew Chen (@andrewtychen) September 22, 2020
We've fallen back into complacent behaviour, Chen says.
But even at our peak of 2.5 million scans a day earlier this month, numbers were still far below the tens of millions of scans that would be generated each day if all adults were using it for every shop, bus and building they entered.
Here's daily scan count data from the NZ COVID Tracer app from 10 August onwards (second wave announcement was on evening of 11 August). Slight dips on weekends. Peaking at 2.5mil scanned on 4-5 September, now trending down with 1.4mil for noon-noon Sunday 20-Monday 21. pic.twitter.com/qOdPcF5lAi
— Andrew Chen (@andrewtychen) September 22, 2020
And now even that modest level of usage has halved.
You could say complacent Kiwis need a kick up the backside, but the problem is global.
Across the ditch, some 7 million of 25 million Australians have downloaded their Government's CovidSafe app (usage numbers are no longer disclosed for "security" reasons).
And even in more compliant Singapore, authorities have struggled to get more than half the population using that country's official app, TraceTogether.
But the city-state is not resting on its hands.
This month it began distributing Bluetooth dongles to all five million of its adult residents. The dongles or plastic tokens can be worn around the neck on a lanyard (like the CovidCard proposed here) or carried in a pocket or bag.
A token exchanges short-distance Bluetooth signals with other nearby tokens. All data is encrypted and logs older than 25 days will be automatically erased. If the user becomes infected with Covid-19 the device will need to be physically handed over to authorities to extract the data needed to do contact tracing.
Local entrepreneur Ian Taylor says Singapore's Bluetooth dongle sounds almost exactly like the initial CovidCard concept he started discussing with Sam Morgan earlier this year.
"It sounds like the concept Sam shared with me back in April, has spent time and resources building, only to be told by people like [Communications] Minister [Kris] Faafoi that he prefers to wait to see what some genius does overseas," Taylor told the Herald.
Morgan subsequently funded development of a prototype CovidCard and helped organise a trial at Nelson Hospital in association with the Nelson-Marlborough District Health Board.
The Government is consulting ahead of a CovidCard trial in Rotorua during October that will involve hundreds.
Morgan says he's folded his team, seeing the trial as a go-slow political sop by a Government that has already decided to put its resources elsewhere. Faafoi has responded that he won't commit "hundred of millions" (Morgan says $100m) of taxpayer funds without a full trial.
Singapore and Australia were among countries who took the halfway step of adding Bluetooth functionality to their tracing apps (it's still being considered for our NZ Covid Tracer).
That means if someone gets infected, a tracer can download data of all their recent close contacts - including those they've forgotten or were never aware of - for much faster tracking. There's no cumbersome scanning of QR code posters required.
There are problems with Bluetooth tracking via smartphone app, however.
Different ages of phones use different versions of the wireless communications standard, and different makes and models of phone have different signal strengths, which can lead to missed connections or false positives. Then you have to cajole people into downloading the app and remembering to open it and leave it open - and allow for the fact that some people (a minority - but in high-risk groups like immigrant workers and the elderly) either don't own a smartphone or do but use it as a dumb phone, never going near apps.
That's why Singapore is going with the dongle - one device, with one Bluetooth standard, with no usability barriers. All you have to be able to do is slip a lanyard over your neck.
And it's why the likes of Taylor want to see a CovidCard here.
There is another solution in the mix: the Exposure Notifications systems co-developed by Apple and Google (Google being the maker of the Android software that runs nearly every phone not made by Apple).
If you opt in, then this system will run in the background on your iPhone or Android 24/7, tracking your every close contact with other Exposure Notifications users.
Apple's iOS 13.7 free software update earlier added an Exposure Notifications options for iPhones here, with our Ministry of Health down as the agency that would issue alerts to your close contacts if you do a positive diagnosis (as with the other systems covered here, your close contacts aren't told your identity, just that they've been in your proximity).
But the software upgrade seemed a premature release. With iOS 14, released this week, mention of the Ministry of Health has been scrubbed. The ministry says it's still accessing the Apple-Google solution, and whether to adopt it - which could involve integrating the technology into NZ Covid Tracer.
Morgan is not a fan of the Apple-Google solution, but he sees the Ministry of Health adopting it over the CovidCard and has now largely checked out of the debate.
Taylor still sees scope for talking the Government around to a CovidCard.
"We are just trying to get someone to sit around a table to examine the problem, from more than simply a medical angle, and see if this is a tool that can help meet the challenge of that problem," he says.
"No one here is trying to sell anything. We just know that this virus doesn't only make people sick - it makes the economy sick as well - and it's doing that faster."
For the Animation Research founder, there could be an element of cultural cringe, with a hardware device not taken seriously until the Singapore Government adopted one.
"The irony for me is this is yet another example of politicians in New Zealand ignoring the innovative thinkers and innovators they have under their noses while they wait to see what someone overseas might do," Taylor says.
For Morgan, the proliferation of technology solutions is a bad thing. The different concepts don't work together, and will only confuse punters.
"Apps have not been shown to change the course of the virus one bit anywhere they have been deployed," he rails.
"Singapore knows this - they went early with an app- and that effective adoption is the key issue.
"Despite this, MoH wants to keep pushing the app barrow - as a form of placebo. perhaps.
"The Apple-Google system is completely disconnected from our manual contact tracers, has massive false positives and massive false negatives, in real-world settings, makes self-isolation completely voluntary, and will not reach the required level of effective adoption anyway.
"Universal hardware, paired with some degree of required usage in places of congregation/ risk has the potential to be a game-changer.
"That's the summary finding of all the work we did with CovidCard. It makes the biggest difference in the scenario where we have a few sporadic outbreaks happening on an ongoing basis, which will quickly exceed the capacity of manual-only contact tracing.
"There are real questions on the ability of apps and wearables to work effectively together and you certainly can't make the Apple-Google system work with dedicated hardware, so the MoH is effectively closing the door on that option with the approach they're taking in addition to just creating more confusion amongst ordinary, non-technical, citizens."
The cost soared from $580m to $924m but output capacity is up on original expectations.