NZ considering $100M contact tracing ‘CovidCard’
newsroom.co.nzKiwi here, I belive this is either snake oil or a conceptual idea rather than a product. The media are touting it as credit card sized device, but with a battery it would need to be much thicker.
Here in lockdown we are allowed to go for a walk or cycle. Potentially this card could register thousands of contacts.
Yeah - I design this sort of thing and I can't understand how the power management of this thing works - battery life has to be a month or so (or it's a cm thick).
Remember that when you're talking at GHz (ie BT) you have to keep your receive PLLs powered up to receive data, that's ~1/2 the power requirements of short distance transmitting
Some media here have claimed a 6 months to a year of battery life. I am very skeptical.
Exactly - I'm not saying it can't be done, just that I'm highly highly skeptical
Just noticed the username. Feel free to shoot me an email, would be keen to hear what kind of work you do in nz.
You don't need to transmit or receive continuously.
If you have two cards that are going to talk to each other one needs to be listening when the other is transmitting, if they're off 95% of the time (like the stuff I design) then 95% of the time someone transmits no one will be listening.
You can try and do something smart around synchonising wakeup times nationwide, but that is genuinely hard (and requires a xtal clock to be awake at all times, also burning power)
I have wondered why Tile here in the U.S. doesn’t do exactly this. Tile “cards” are about the thickness of a credit card. I have one in my wallet so I can “beep” my wallet when I can’t find it.
The power needs must be miniscule, I assume.
Anyway I think its technical feasability is established vis a vis powering electronics in that form factor.
And gathering a lot of recorded contacts is exactly what the goal is, so that’s not a bad thing.
Yes, it would generate a huge amount of data to sift through.
Based on what we know about the proposal, the hardware platform doesn't yet exist. Even if the full design was already complete, how long would it take to tool up a factory to produce 5,000,000 units and the distribution infrastructure to ensure everyone had a unit? It just doesn't make any sense.
There's been a disturbing trend where businesses are approaching government agencies (without even an RFP being issued) trying to get them to endorse/fund their app or device or technology. It just stinks of an attempt to profit from the disaster as opposed to a genuine desire to help enhance our contact tracing infrastructure.
Surely just installing an app voluntarily would get enough tracing to be effective?
As a kiwi who has spent half my life living in different countries, I can say trust in the government here, while not universal, is higher than anywhere else I've been. I think most people would do it
Agree. compared to the shit-storms politicians are creating in other countries I'm sort of proud of the way our government acted so decisively and unilaterally, and I generally get the feeling they are acting for the good of the people not themselves.
I can already see the cracks appearing as some idiots try to score points or promote their own agendas and the media are happy to give them air time, the "covidcard" or phone tracing app will open another one of those cracks even though like you say the majority would conform
I trust our government quite a bit too. Would have no qualms downloading and using an app. I want an app. I'm less keen on the card, but honestly whatever gets the highest ability to trace the most number of contacts is what I'm for. The numbers out of Singapore dont seem to indicate an app alone is a good enough solution.
The hybrid approach the article suggest where perhaps the card + app could be compatible would be ideal.
i trust the govt, but would not do it. a self reporting system is better than giving up privacy rights.
"The only central register of card IDs would be held by the Government and isolated from other Government databases, so it could not be used for anything other than Covid-19 contact tracing."
What does "could not" mean, anyway?
Same argument currently being brought by the German PEPP-PT project. I really don't get why they push for these centralised solutions when clearly defined decentralised models exist.
>I really don't get why they push for these centralised solutions when clearly defined decentralised models exist.
power
The big problem with all radio based contact tracing solutions is the receiver. Receivers use at a minimum several mW of power. A typical Bluetooth low energy receiver will use about 50mW so will drain a coin cell battery in a day. Bluetooth low energy saves energy by transmitting for very short periods of time so transmit power is low but receiver power can be high. Even asking most consumers to leave their phone's Bluetooth recovery on all day is a big ask because of the impact on battery life.
Schemes could be designed to better synchronise receivers and transmitters so the receiver can be switched off most of the time but this currently can't be done with Bluetooth.
Your phone's Bluetooth scanning IS always on if it's an iPhone or Android. Always. By default.
Yes, and it has a honking big battery that you charge every day or so.
A card with a proposed 6 month battery life with receiver PLLs that are always spun up probably needs more battery than you can fit in a tiny thin card
That may be true when the screen is on, but from my experience scanning is not switched on in sleep modes on iPhones and most Android phones. I don't have definitive proof of this but I have never seen a modern phone be able to reliably scan whilst in sleep mode for any long period of time.
it is on just at a lower duty cycle.
source: worked on this for a few years at googl, both chromeos and android
Once again we ignore Asia and their clear lessons with covid-19 because they are not enough like us.
Electronic solutions don't work, even in compliant countries like Singapore they don't, it's boots on the ground that does.
Here's a good write up on Australia's attempts as they stand -
https://medium.com/swlh/some-basic-points-on-contact-tracing...
With all the issues with tracing apps listed. I can see the systems around the world causing more problems than benefits for the person in the street.
Possibly the concern for many of ending up causing massive issues in their social and work circles. Will mean many simply wrap in tinfoil if indeed they bother to carry at all.
No body wants to be known as the person who sat the otherside of a drywall partition to someone covid 19 positive. Which meant they then basically quartined for two weeks all their work mates in a role which was not WFH compatible and hit them all in the wallet.
The arguments the promoting organization is using against a smartphone app (e.g. Google and Apple's thing) are pretty dubious. That seems like the easiest way to do digital tracing by far.
"Without fast case isolation after restrictions ease, modelling shows that the virus could spread exponentially and ultimately kill 100,000 Kiwis."
The left out the word "unreasonable". "Unresaonable modelling shows..."
See the Tailrisk analysis: http://www.tailrisk.co.nz/documents/Corona.pdf
Yeah but the Tailrisk guy is not an epidemiologist, he's out of treasury and naturally has is more economy focused outlook.
One person's "unreasonable" can be another's "sensible caution"
You're right about that guy. So I'll be more specific. The model that shows large numbers like 14,400 (not even 100,000) presumes we do NO contact tracing at all (not just slower contact tracing). I'd wager most people would agree that is an unreasonable assumption.
That analysis is a perfectly good 'model', but it makes a few interesting/weird choices and assumptions of its own.
"The true burden of the epidemic can be calculated by applying an factor of around 0.15 to the number of account for life years lost. 500 deaths becomes, 75 on an adjusted basis, and can be compared with the 350 lives lost on the roads each year" - He means, of course, the true economic burden, which is a specifically different thing.
Putting that aside (although it clearly reflects his economic bias), actually implementing his model assumes that we can request the public to maintain a variety of distinct contact levels accurately, Im not convinced that we could expect the public to do that effectively over time.
His model is also extremely dependent on accurate testing and contact tracing, which grows exponentially harder as we increase the level of allowable contact.
His model is fine as far as models go, but it definitely makes the assumption that things like level of contact and contact tracing are as easy to adjust and manipulate as his normal treasury numbers, and I personally believe that is a fallacy.
I think if they could both work in conjunction with each other that would be ideal. I'd try the app for sure, but success rate doesn't seem to be high enough in Singapore for an app only to be the sole solution. Use as many approaches as you can make compatible ups the chances of uptake.
This reminds me of schools early on who wanted to give all their students ipads.
Hello? fundamentals go much further than tech magic bullets.
For instance, the chinese took people's temperatures.
The conspiracy theorists almost had it right. It’s not tinfoil hats we need, it’s tinfoil bags and pockets.