Well that aged well
Centralised app now abandoned in favour of Apple/Google decentralised app.
Once described as a key part of England's COVID-19 test-and-trace system, the smartphone app being trialled on the Isle of Wight is no longer a priority for the UK government and won't be ready until winter. In May, Prime Minister Boris Johnson said England would have a "world-beating" tracing system from June, the Department …
My guess is the delay until Winter is because their centralised app probably didn't work as expected., so now they have to go back and start over with a new one based on the Google/Apple APIs.
The danger with waiting until winter is that will there be enough up take of it? as there is already some COVID19 fatigue where people have started to switch off from the news because they are fed up about everything being about the virus.
I think we're both right.
According to Google, their API does more than you suggest: "The Exposure Notification Bluetooth Specification ... uses Bluetooth beaconing to detect proximity." - so an app developer doesn't have to solve that problem, they just need to turn it on.
However, what it does is a bit less than I thought. But Google do provide sample app code and sample notification server code - see https://www.google.com/covid19/exposurenotifications/ - so anyone could build an Android prototype on the back of this very quickly (but won't be able to publish the app).
So I still think the hardest low level technical issues have been taken away from the app developer
"But Google do provide sample app code and sample notification server code - see https://www.google.com/covid19/exposurenotifications/ "
I saw this in action using ADB on my ancient Android phone I bought many years ago.
This cheap phone never received any firmware updates but I could see that the com.google.android.gms app had a new permission added of:
Gov claims to already have both apps. Their version doesn't work well with iPhones, and the Google/Apple app doesn't do a good job of estimating distance. They hope to persuade Google and Apple to adopt their distance-estimating code. Part of the delay may be waiting for that to happen.
Good distance detection is important to avoid false positives. People will have to self-isolate for 2 weeks based on this app, so any positive is disruptive. Gov says they won't release it until the false positive rate is low enough. (Part of the goal of the centralised app was to collect enough information to help trackers reduce false positives.)
This was from the daily briefing.
This post has been deleted by its author
"COVID19 fatigue where people have started to switch off from the news"
I couldn't get enough information about it for the first month or two and watched the daily briefing religiously every day. Now I'm frankly so bored by the whole thing that I haven't watched the briefing for a couple of weeks and change channel when the news interviews politicians about the subject.
The BBC and Sky news to a lesser extent have also become extremely tedious, picking random families across the land and interviewing them on "how they feel" about this or that related to the lockdown etc.
I know I *should* keep informed on developments and the latest policies, but my boredom threshold has been breached and I suspect that of many other people has too.
I think that getting folks to install the tracing app come Winter may be an uphill struggle.
The general news is not a good source because it focuses mainly on the numbers and then the "human interest". There's so much other stuff going on, but this is how the news cycle works and fatigue is normal.
For example, German news yesterday decided to report that Sweden now has > 5000 deaths and a mortality rate five times that of Germany's. It didn't mention that reported cases have increased recently as Sweden is now testing more. Nor that admittances to ICU wards and fatalities continue on a downwards trend. It also claimed that Sweden's approach is controversial but failed to mention that the main party of government continues to rise in the polls. All policies are detabable but FWIW in Sweden the main difference has been not closing schools and nurseries or restaurants. But people have still been asked not to travel even in Sweden and hospitals and care homes have strict hygiene regimes.
The focus on numbers keeps people scared, makes whipping boys easy to find but doesn't explain much. It also doesn't really prepare us for a possible spike in the autumn: we're not going to eradicate infections before then and we still don't really know what the best protective measures are. Oh, and along wih age, socioeconomic status is the biggest determinant for morbidity and mortality. But let's not talk about that because poverty implies class and discussions about prevent the middle-class from feeling self-righteous.
Try and listen to a general science report if you can as this should contain less repetition and more information. With so many resources being poured into research there is some very interesting stuff coming out and cooperation worldwide is impressive. But there is also an awful lot of duplicated research based on undersized studies, which means opportunities lost.
I'm generally a big fan of the news but this focus on quantitative reporting and the "won't someone think of the children" outrage as justification for removing people's freedom has made me increasingly sceptical: I know what the news will be before it's broadcast.
It looks increasingly like keeping schools open was an entirely legitimate thing to do. "What we’re seeing more and more from the data that comes out is that child-to-child or child-to-adult spread is actually not common”. https://www.the-scientist.com/news-opinion/covid-19-is-very-different-in-young-kids-versus-adults-67637
Sweden has the most data on this and it does indeed seem that closing schools had little effect on the spread. Not that they don't carry the virus, but they're not the biggest risk. Which is why neighbouring countries quickly reopened schools once their peaks had passed. We're now also seeing data on how keeping kids out of school has affected their learning and it's not good.
Doesn't stop the usual kneejerk reactions though: schools and nurseries have been closed in Gütersloh in Germany due to infections in a local meat processing plant, where the majority of the workers are from south east Europe and who live without families in what are basically barrracks.
Science, we've heard of it.
I gave up listening to the briefing about a week in when it became clear that there was no useful information being given by the government - just a load of "we're working really hard", "it is a top priority", "we're laser focused", "we're following the science (but not going to tell you what the science is)". It was all just a load of waffle and platitudes.
I fear the reality is that the govt actually did sweet FA and had no idea what was going on, it was just the various agencies getting on with things on their own.
And then our 'world beating' app being boasted about so Boris could make his beloved brexiteers feel all superior to the EU. All they did was line the pockets of outsources for delivering stuff all. God, next thing you know they'll be giving shipping contracts to companies that don't have boats !!!
Yes, it is available as open source and, because it uses the Google/Apple API compatible in theory at least with similar apps in other country, though apparently no one has actually done any work on interoperability.
Anyway, now that it's finally available the news has stopped reporting on it as much. Which is good. In a couple of months we'll know more about its effectiveness. More important was the decision that anyone who wants a test can have one and the health insurance will always pay for it.
The danger with ANY app is whether enough people will use it. Be it centralised or decentralised most of the public do not care. Some will use it many won't and a significant proportion can't.
Given that 25% of contact tracing attempts fail as well I am unclear how any App is going to make a significant impact. You are correct in that there is COVID19 and lockdown fatigue but we also live in a society where a large enough number of people simply don't give a stuff.
Look at the crowds on beaches, queues for shopping and the demonstrations.
Ironically I expect many of these people also have kids that they will not send to school because it is "not safe".
"They won't take you seriously at that price"
Regrettably true. They're happy paying 900k to "rebrand" a plane. And the BoE has just given them 100 billion to spend.
Meanwhile, at Crapita HQ the business development team are typing up the latest "as suggested by el reg badge holders" proposal...
I think the new £100bn in QE is for commercial banks to use for loans to assist the economy where the COVID-19 specific schemes aren't applicable. BoE has already extended the govt's contingency fund (its overdraft facility, if you will), which is different to normal government borrowing.
The German one cost a mere € 20 million. Mind you, it was developed by T-Systems and SAP so what do you expect. As to whether it works: only time will really tell. They might help a bit.
But these apps can never replace testing. The faster you can test people, the less contact-tracing you have to do. But testing is more expensive…
They aren't supposed to replace testing. The whole point is that they make life easier / the process quicker once you have been tested positive.
It is an aid for the people who have to do the contract tracing after a positive test result. Or rather, in the German app's case, the user captures the QR-code they get with the test results and upload the contact information they have and other users of the app who have come into contact with them (<2M for >10 min.) will be notified that someone in their vicinity was tested positive within the last 14 days and it is recommended that they get themselves tested as well.
You don't know who was infected, where or when, just that you had contact with a verified case of Corona. It is then up to you, whether you take it seriously or not.
You don't have to use the app, and you don't have to get yourself tested or even check to see if anyone you "met" was infected. That is all optional.
The whole point is that they make life easier / the process quicker once you have been tested positive.
It has no effect for anyone tested positive. What it might do is help track the spread in other people with anyone who may have been in contact. That's the theory, we'll find out over the next few months as to whether this is much more useful than the existing approaches, which after a poor start, have proved very effective in Germany.
Personally, I'm more inclined to believe that the registration of people in restaurants, et al. may prove more effective because it provides the locus. But nothing has been done to digitise this. But time will tell.
It has no effect for anyone tested positive.
It was never supposed to. It is there to help others who may have come into contact with the infected person.
But, I agree with the rest of your post completely. Looking at the rest of the world, we don't seem to be doing too badly - let's just hope there aren't too many Westfleisch and Tönnies out there, waiting to whop the statistics in the arse.
I have an image of Clemmens Tönnies sitting in the boardroom laughing as the Westfleisch story broke, then not actually doing anything, until the smile was wiped from his face this week...
Having worked in the meat industry for around a decade (software supplier to slaughter houses and meat processig plants), I've seen the conditions in small and large slaughterhouses (but not Tönnies). I know a couple of "honest" companies that used local workers, but the pressure put on them by the discounters to push down prices and the shareholders to turn a profit means that most of the big ones have resorted to "importing" cheap labour out of the former East Block and putting them up in rundown flats, where they share beds (the workers on shift swap places with those going off shift) in many cases.
We have become so accustomed to cheap meat that many are no longer willing to pay for quality meat and/or eat less meat, like we used to. As a child, I grew up in a middle-class family and we had meat probably 4 - 5 days a week, although that was often a roast on Sunday and leftovers the next day or two. My wife's family had meat 1 - 2 days a week. Now, even poor families have meat most days of the week. But the quality of the meat has sunk dramatically and the price is unrealistically low, in Germany.
I assumed "winter is coming" was a dig at Raab saying 'taking a knee' came from Game of Thrones.
This latest u-turn hardly inspires confidence in the government, but I guess I'll install it. It's sad that I trust Google-Apple on privacy more than my nation.
ETA: I just heard it will still be a government app but based on the Google-Apple model. So no ta.
My parents are old enough and frail enough that I really do care. Yet the process around getting these apps going seems to be one of never-ending debate over privacy issues. As the first wave of infection has declined complacency seems to be setting in and I see little sense of urgency. My reading of the stats suggests the chances are there will be a second wave.
We clearly need to be pushing ahead with both threads of development - the one using the Google/Apple API that seems to protect privacy while largely precluding epidemiological research, AND the independent alternative that seems to run up against issues with iOS. That this second option has proved so problematic for so many countries seems to me to be a huge indictment of how the industry has evolved over the past decade.
To any developers out there: please just get ***** thing working.
Epidemiological research is already possible, see the app from Kings for example. Trying to make it identical to contact tracing is a problem. You do not need to recruit everyone in the country to do research, and this is the barrier they have unnecessarily raised for themselves.
Have you a reference for this Kings app? My guess is that the work that is achievable under the heading of epidemiology will be very severely constrained. It will not, I suspect, be able to work with the wealth of location and profile data that I naturally assume Google is already holding on me and every other Android user.
Remember. Real people are dying. Just get the ***** thing working!
They've already published some work I believe.
I should add that a contact tracing app will not simply make coronavirus go away. Though it will possibly pick up contacts that manual tracing does not (and miss some it does) https://www.lshtm.ac.uk/newsevents/news/2020/uk-modelling-study-finds-case-isolation-and-contact-tracing-vital-covid-19, meaning the "we find people prefer the personal touch" answer given is misinformed about the difference between the two approaches. Whether it's intentional to confuse the two or not I don't know. (Manual contact tracing also relies on a website, and apparently people are encouraged to fill their contacts in online, so the personal touch is somewhat missing there.)
The app is at https://covid.joinzoe.com/. I've been using it for several months.
I understand some of your anxiety, but please remember that *across the world* fewer than 500k people have died of this virus (I'm taking the current figure, since the claims of under- and over-reporting seem to balance out at the moment). This is not an "at any costs" situation by any means - let's look at the broader picture.
I don't entirely know about balance out, excess deaths numbers for some South American countries seem to suggest that coronavirus deaths are essentially not being reported. In Ecuador they've started distributing cardboard coffins for free and that doesn't feel like it tallies with a few dozen deaths a day in a country of 17 million. (Of course, maybe it's a publicity stunt, or maybe they always use cardboard coffins.) But that's a country that is not going to be helped by a UK tracing app anyway.
I've had to upvote because I don't feel you deserve a downvote just for asking the question.
However, on real people are dying. Yes, although as Intractable Potsherd points out it is still relatively limited. That's not to say it should be ignored, hundreds of thousands of people are dying before they should, tens of thousands in this country alone. And we should get the tracking app working, there's no reason but arrogance that we have spent so long fannying around rather than using the solution that two of the largest technology countries in the world have developed to work on their platforms. But the app is only going to make a degree of difference, and throwing out rules on privacy, creating something that's half-baked just for the sake of creating something, those are not good ideas.
My area of research is dementia. It's responsible for over 10% of deaths in the UK https://www.alzheimers.org.uk/blog/research-dementia-UK-biggest-killer-on-the-rise not just for a few months but for years and it's increasing. This is important. But I wouldn't advocate forcibly recruiting people to studies or throwing away their right to consent or not to have their data used for research, it undermines the trust of the people whose help we need for our work, contradicts our concern for their wellbeing, and in the long run harms us all.
I've never before felt that literally everything my Government says will turn out to have been actual bollocks. About the only thing keeping the infection rate down at the moment is the fact that it's raining, and I've no doubt they'll find a way to take credit for that.
I think it was meaningless from the get-go. Just another silly system involving five things to try and give the illusion of there being a plan. Perversely enough it's actually given Whitty a tool to express his displeasure, though I'd rather he just came out and slagged them off properly at one of the briefings…but I presume he's disinclined to do that because he knows he'd be replaced immediately. UK governments do not seem to like evidence-based anything.
So after years of ignoring or putting the phone down on cold calling call centres trying to spam us with crap or other scams, the Govt now seriously expects people to start answering the phone to random callers who are telling them they have been in contact with someone infected?
Shirley it doesn't take a genius to work out how well that will go down, or indeed how long it will take for all of these scam peddlers to start the impersonation game and use this as yet another way to hook the gullible and vulnerable (given that iirc it's actually started already even before the contact tracing has)?
Would you trust a cold call on such a subject?
No. Especially as they are not allowed to say who you may have caught the virus from for data protection reasons; so you have no anchor on reality to substantiate their claim.
I'd hear them out though, but the moment they want either financial information or to "take me through security" I'd hang up on them.
It's all Apple's fault, apparently.
But we all knew that before HM Govm't spent a load of money trying to develop something that they knew would not work in the Apple walled garden.
But hey why listen to the technologists when BoJo has spoken (probably with Cummings's arm pulling his vocal chords.}
It's clear there are now two routes:
1) Try to develop a centralized app. A centralized app would allow contact tracers to interact with it, it would allow infection spread to be monitored so localized lockdowns could be initiated more easily. However, privacy groups think it could be misused (and it probably would be) and Google and Apple think they know better than democratically elected governments. (Tony Benn's five tests don't work too well for Apple and Google's apparent control over governments' response to this situation.) The app is a failure because it's trying to bypass in-built security systems.
2) Use the Google/Apple model. Since this is decentralized, it is of effectively zero use unless the app proliferates through a large section of the population, at least 60% is the mooted number. As it is privacy focused, it cannot interact with human contract tracers. Only the person who is infected knows, and can choose whether or not to share this information with anyone else. No enforcement of quarantine is possible, no tracking of the epidemic is possible, everyone is in the dark. And it's not clear that the G/A API works either. Under the app, there will be two parallel systems of contract tracing, from the app and from humans.
So this is probably why the app is on the back burner. It's not going to help in either case.
The centralised version is also no use for contact tracing unless a large enough proportion of the population use it. And whatever you build on top of the google/apple model can interact with contact tracers, just not in detail about who your contacts were, but it could put you in touch as 'exposed' and get you scheduled for a test. The promise of not just having to hole-up for a fortnight unnecessarily and having a positive test result if you did need to I think would really increase compliance. Bear in mind the 1/4 not 'engaging' with our current contact tracing process, some false negatives would be worth bringing that number down (it'd also be interesting to see data about the infectivity of false negative people, because I'll bet that means lower virus levels in the airways and less transmission, hard to measure though for obvious reasons).
"The centralised version is also no use for contact tracing unless a large enough proportion of the population use it."
I was thinking more of "John has the virus and the app, Steve and Fred have the app as well, so they are already notified. Just have to deal with everyone else." That's how it interacts with the app, by stopping doubling up the workload.
"The promise of not just having to hole-up for a fortnight unnecessarily and having a positive test result if you did need to I think would really increase compliance."
Nah. Current guidelines are:
1) Hole up for 14 days regardless of a negative test, and
2) No tests for asymptomatic people.
So the guidelines are the main problem. Nobody will do it.
I’ve been listening to a lot more old radio output since lockdown. 40 years on and Yes Minister is just as relevant as it was then. Except now we have expensive outsourcing added to the bloated and inefficient Civil Service. Nothing significant ever gets achieved and the public are fed self-serving nonsense by politicians.
Biting the hand that feeds IT © 1998–2020