APHL partners with Apple, Google and Microsoft on national COVID-19 Exposure Notification ...

Posted:
in General Discussion edited July 2020
The Association of Public Health Laboratories is working with Apple, Google and Microsoft to build a national server that will securely store COVID-19 Exposure Notification data, a system that promises to bolster state and regional efforts to contain the virus.

Exposure Notification


Detailed in a blog post Friday, the project aims to develop a more comprehensive and cohesive exposure notification solution for the U.S.

Instead of storing user contact tracing keys -- critical, time-sensitive information -- on multiple, unlinked servers run by state agencies, APHL offers to securely compile and make that information available on a national server. Storing keys of affected users on a single database eliminates duplication and can enable notifications across state borders, the group said. Further, states and territorial agencies that integrate with APHL's proposed server would be able to more rapidly build out exposure notification apps.

The system implements Apple and Google's Exposure Notification API, which uses random device identifiers -- keys -- to generate temporary IDs that are sent between devices via close proximity Bluetooth communications. By swapping keys, apps integrating the Apple-Google system can track and notify users when they are exposed to others who test positive for coronavirus.

With security at the fore, the solution does not store data on central servers run by Apple or Google, but instead silos anonymized Bluetooth beacons on user devices until participants elect to share the information with an outside party. If and when a user is diagnosed with COVID-19, they can opt to upload a 14-day list of recent anonymized contacts to a distribution server, which matches beacon IDs and sends out notifications alerting those individuals that they came in close contact with a carrier of the virus. Doctors can also peruse the data, if such access is granted.

Those protections extend to the APHL initiative.

While not specifically outlined in today's announcement, APHL would likely take over the role of providing a contact distribution server, in this case for the entirety of America. Microsoft is partnering on the project to provide the national cloud-based key server based on an open source design created by Google Cloud.

"We're honored to partner with Apple, Google and Microsoft to make this groundbreaking technology accessible to state and territorial public health agencies," said Bill Whitmar, president of APHL and director of the Missouri State Public Health Laboratory. "Apps using this technology will rapidly inform users of a potential exposure to COVID-19 and provide them information they can use to protect themselves and their families."

APHL has extensive experience in connecting public health laboratories, entities and government agencies. The group's Public Health Laboratory Interoperability Project launched in 2006 as one of the first systems to allow for the exchange of standardized data between public health entities. That was followed by APHL Informatics Messaging Services, which has evolved from a one-way conveyance of critical health data to a cloud-based platform that "transports, translates, validates and hosts data for federal, state and local public health agencies," according to the post.

Exposure notification technology shows promise, though adoption needs to reach critical mass in any given population before it becomes an effective tool in staunching the coronavirus tide. So far, only a handful of countries have launched apps that integrate the API. No U.S. states have followed suit, though three -- Alabama, North Dakota and South Carolina -- have signaled intent to do so amid a resurgence in cases across the country.

Comments

  • Reply 1 of 16
    JWSCJWSC Posts: 1,203member
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.
  • Reply 2 of 16
    mknelsonmknelson Posts: 1,125member
    JWSC said:
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.
    Polls have shown a wide range of interest. The US may be right out as you suggest, other (hello from Canada) countries seem to have a better understanding/less politicization  of the benefits of contact tracing (it's not tracking).

    Having said that, I have a few co-workers who had a knee jerk reaction when they saw the API in the iOS update last month. They clearly didn't understand the methodology or privacy protections built in to the system.
    edited July 2020 Rayz2016JWSCjony0
  • Reply 3 of 16
    russwrussw Posts: 21member
    JWSC said:
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.
    I hope you’re wrong but fear you’re right about the US. So funny when you realize your cell company is tracking you already, not to mention Google and Facebook. 

    Agree at least this will be a trial run for the next pandemic and let’s hope there is more reason to trust government at that point. 
    Rayz2016JWSCjony0
  • Reply 4 of 16
    DAalsethDAalseth Posts: 2,783member
    If Canada has an App based on the Google/Apple APIs that works the way they are supposed to, then yes. I would install it. that keeps the data secure from after the face spying or data theft. All notifications were to go from device to device.
    This idea to have a centralized repository though...no. It violates the core principle of the G/A system. The whole essence of it is to be secure because there is no data on a central server that WILL be hacked, or receive a warrant to trace people. 
    So yes to the app.
    No the this though.
    edited July 2020 anantksundaram
  • Reply 5 of 16
    sdbryansdbryan Posts: 351member
    Finally some good news about doing more than just trying to avoid Covid-19. I agree that in the short run there are not good prospects for sufficient adoption but given present leadership we sadly may have plenty of time for adoption to grow while the threat continues to be a fact of life (in the US).

    As indicated by previous individuals, anyone with a cellphone is continually sharing identity and position in order for the phone to function. This protocol shares neither identity nor position. It is just about notification of possible exposure and thus is much more privacy protective than the billions of cellphones in use.
    gatorguylarryjwgregoriusm
  • Reply 6 of 16
    anantksundaramanantksundaram Posts: 20,404member
    sdbryan said:
    Finally some good news about doing more than just trying to avoid Covid-19. I agree that in the short run there are not good prospects for sufficient adoption but given present leadership we sadly may have plenty of time for adoption to grow while the threat continues to be a fact of life (in the US).

    As indicated by previous individuals, anyone with a cellphone is continually sharing identity and position in order for the phone to function. This protocol shares neither identity nor position. It is just about notification of possible exposure and thus is much more privacy protective than the billions of cellphones in use.
    To blame it on the "leadership" makes no sense. I doubt that those who are anti-mask or anti-tracing would be any different regardless of who the "leader" was.

    That's just the US, warts and all. It is certainly not a place for everyone, not a place where people just fall in line, it's what it is.
    williamlondon
  • Reply 7 of 16
    BeatsBeats Posts: 3,073member
    Stinks of data mining and exploitation.

    This is the fragmentation you Apple haters wanted. Now you got it.
    williamlondon
  • Reply 8 of 16
    MarvinMarvin Posts: 15,322moderator
    sdbryan said:
    Finally some good news about doing more than just trying to avoid Covid-19.
    That will still be the best option for now. Tracing people will flag possible exposure but when people get a notification, isolation is the only solution. It's not clear how much children will influence things. They'll likely be traced less, possibly not at all, and it's pretty hard to do in a school setting. Some reports say kids can spread it, others say they don't:

    https://www.msn.com/en-us/news/us/can-kids-spread-the-coronavirus-conclusively-without-a-doubt-e2-80-93-yes-experts-say/ar-BB16QSYT
    https://www.bloomberg.com/news/articles/2020-06-23/school-children-don-t-spread-coronavirus-french-study-shows

    Tracing people can help with easing the lockdowns as it could flag an outbreak and allow locking down regions instead of countries but it needs active participation to be effective and for people to update their mobile devices. The slowest part for tracing will be getting a test to confirm infection. There was a trial of a tracing app that had positive uptake:

    https://www.dailymail.co.uk/news/article-8525395/Test-Trace-DID-work-Isle-Wight-study-finds.html

    It was on an island, which is easier to control and there were other factors involved but it must have helped with some cases. They mentioned not being able to track distance very well. Bluetooth has a range of up to 100 metres and some apps warn people if they've come within 100m of anywhere with an infection:

    https://qz.com/1810651/south-koreans-are-using-smartphone-apps-to-avoid-coronavirus/

    I guess it's safer to be overly cautious but trains at a station with 300+ people each would be in proximity of someone on the platform who might never get on the same train. The documents say there are measures available for exposure time and signal attenuation:

    https://www.apple.com/covid19/contacttracing/

    The app developers will need to do some tuning to get the best results. Since millions are already infected, that's a lot of already untraced contact points. It's worth trying everything to help ease the spread and to get more information about how and where it spreads but it ultimately needs immunity. When the cold/flu season starts, people aren't going to know what they have.
    DAalseth said:
    This idea to have a centralized repository though...no. It violates the core principle of the G/A system. The whole essence of it is to be secure because there is no data on a central server that WILL be hacked, or receive a warrant to trace people. 
    People use centralized services all the time. iTunes/App Store/iCloud/Apple TV/Google Play/GMail/Amazon/eBay are all centralized services. What's important is what they are storing about you, not that they are storing information. A central server needs to be used for notifications, Apple and Google describe how people need to set it up:

    https://developer.apple.com/documentation/exposurenotification/setting_up_an_exposure_notification_server
    https://github.com/google/exposure-notifications-server

    Devices don't send out notifications peer-to-peer, the apps routinely query the notification server for reported infections for matches with the device's keys. All tracing methods require some compromises to user privacy as the notification server can track a request IP and the apps can report the match to the server but a random app from the App Store can do worse than this (Safari's reader feature can let you bypass the paywall if it pops up):

    https://www.nytimes.com/interactive/2018/12/10/business/location-data-privacy-apps.html
    dewme
  • Reply 9 of 16
    jony0jony0 Posts: 378member
    DAalseth said:
    If Canada has an App based on the Google/Apple APIs that works the way they are supposed to, then yes. I would install it. that keeps the data secure from after the face spying or data theft. All notifications were to go from device to device.
    This idea to have a centralized repository though...no. It violates the core principle of the G/A system. The whole essence of it is to be secure because there is no data on a central server that WILL be hacked, or receive a warrant to trace people. 
    So yes to the app.
    No the this though.
    I would like it to be centralized here as well instead of having each province have their database and I think that's what they intend to do after testing in Ontario only. This concept is a good thing. Remember it's only anonymized tokens that have been sent voluntarily by an infected person.
    Also remember that the core principle of the AGEN was to not store every token and particularly no location. This does neither and in fact they wanted to restrict it to one app per country anyway, excluding the US. So this falls in line with the original principle.
  • Reply 10 of 16
    seanjseanj Posts: 318member
    The irony is that this implements the functionality that those countries who tried to build a centralised tracking app (eg France, UK, Germany originally) originally wanted. This gives authorities the ability to identify not individuals but local outbreaks, allowing them to target additional testing, medical resources, etc to specific cities or even neighbourhoods.
  • Reply 11 of 16
    larryjwlarryjw Posts: 1,031member
    In the UK, they have 336,000 regular users of Kings College of London app as part of the COVID Symptom Study. That's a whopping N for a study with substantial power to find symptoms which have been potentially missed.

    They've discovered that 8.8% who tested positive had experienced a skin rash, compared to 5.4% who tested negative. Similarly, 8.2% of users who tested negative but also had Covid symptoms reported having a rash. A followup study of 12,000 showed 17% testing positive had a rash as the first symptom, and 21% who tested positive had a rash as the only symptom. 

    Thus, the UK study seems to have decently confirmed that a rash is one of the symptoms of Covid.

    This illustrates the value using apps. 
  • Reply 12 of 16
    gregoriusmgregoriusm Posts: 513member
    Took little, too late? If it saves one person, it ain’t too little, too late. 
  • Reply 13 of 16
    GeorgeBMacGeorgeBMac Posts: 11,421member
    APHL?
    Isn't that the organization of the labs who completely blew Corona virus testing in the U.S.?
    The organization of private and public labs who still take a week to return test results?
    That organization?

    Why would I trust them to extend their dominance?
    They should maybe first fix their own existing problems before creating new ones.
  • Reply 14 of 16
    mike1mike1 Posts: 3,284member
    JWSC said:
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.

    Except round 2 of this is probably around the corner. Let's get through the fall and winter.

  • Reply 15 of 16
    GeorgeBMacGeorgeBMac Posts: 11,421member
    mike1 said:
    JWSC said:
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.

    Except round 2 of this is probably around the corner. Let's get through the fall and winter.


    That part could be tricky.
    As Einstein advised:   insanity is doing the same thing and expecting different results.

    Originally, in the U.S. there were only a handful of infectious people roaming the streets who, ultimately,  caused the infection of 3 million.  Why?  because we had no defense.   The best we came up with was to huddle in our homes.   But that was a temporary solution.

    So, come the fall, we will still have infectious people roaming the streets and our only defense, huddling in our homes, has been declared done and over with.   The good part is that more people are wearing masks and that will help, but essentially, we have not made much progress:   So, the virus is still here, and not much has changed -- so why would we expect different results?
  • Reply 16 of 16
    mike1mike1 Posts: 3,284member
    mike1 said:
    JWSC said:
    I suppose everyone who takes part in these COVID-19 tracking solutions feels good about this work. And I’m sure it’s good work. But getting it deployed to the masses, many of whom will not voluntarily submit to tracking, is just too little and too late.

    It may be a prepackaged solution for the next pandemic though.

    Except round 2 of this is probably around the corner. Let's get through the fall and winter.


    That part could be tricky.
    As Einstein advised:   insanity is doing the same thing and expecting different results.

    Originally, in the U.S. there were only a handful of infectious people roaming the streets who, ultimately,  caused the infection of 3 million.  Why?  because we had no defense.   The best we came up with was to huddle in our homes.   But that was a temporary solution.

    So, come the fall, we will still have infectious people roaming the streets and our only defense, huddling in our homes, has been declared done and over with.   The good part is that more people are wearing masks and that will help, but essentially, we have not made much progress:   So, the virus is still here, and not much has changed -- so why would we expect different results?
    Exactly. Which is why I was commenting on the original poster's comment regarding it being too little, too late. It could help with round 2.

Sign In or Register to comment.