[LINK] America Is Reopening. Coronavirus Tracing Apps Aren’t Ready.

[Bernard Robertson-Dunn]

[Following on The IT debacle that could cost lives

[https://www.themandarin.com.au/135285-opinion-covidfail-the-it-debacle-that-could-cost-lives/

[Who knew IT could be so hard?

[At least we now know Australian bureaucrats and technologists are as
good (or bad) as the rest of the world

 America Is Reopening. Coronavirus Tracing Apps Aren’t Ready.

Wall Street Journal

https://www.wsj.com/articles/america-is-reopening-coronavirus-tracking-apps-arent-ready-11592845646

22 June 2020

Smartphone apps meant to track where people have traveled or whom they
have been near are mostly buggy, little-used or not ready for major
rollouts, raising concerns as restrictions lift and infections rise

Local officials in Teton County, Wyo., home to Yellowstone National Park
and resort town Jackson Hole, want to prevent a new wave of coronavirus
cases as the area reopens. They decided to lean on technology.

The county signed up for a location-tracking app developed at the
Massachusetts Institute of Technology
<https://www.wsj.com/articles/mit-researchers-launch-location-tracking-effort-for-the-new-coronavirus-11585315674>to
help accelerate contact tracing, the process of notifying and isolating
people who might have been exposed to the virus.

But as tourists stream into Yellowstone—rangers spotted license plates
from 41 states the day it reopened in mid-May—the app isn’t ready. It
can’t accurately track location, it’s missing key features and its
developers have struggled to protect sensitive user data.

U.S. states and counties are placing great faith in contact tracing, in
tandem with aggressive testing, as they reopen their economies. Pressure
has increased as coronavirus infections rise in many states, including
Arizona, Texas and Florida.

The quick spread of the coronavirus makes it hard for human contact
tracers to keep up, so authorities are turning to smartphone
technologies to help track where people have traveled or whom they have
been near.

What is emerging across the country so far, however, is a patchwork of
buggy or little-used apps
<https://www.wsj.com/articles/curbing-coronavirus-with-a-contact-tracing-app-its-not-so-simple-11588996809>,
made by partners ranging from startups on shoestring budgets to
academics to consulting firms. Some are working with location-tracking
firms that have been under fire from privacy
<https://www.wsj.com/articles/once-pariahs-location-tracking-firms-pitch-themselves-as-covid-sleuths-11592236894>
advocates.

None appears ready for a major rollout, even as more local governments
ease restrictions.

Utah signed a deal worth more than $6 million with a firm backed by the
family of billionaire Nelson Peltz and other investors. Rhode Island
hired Indian software company Infosys <https://quotes.wsj.com/INFY> Ltd.
to build its app free. North Dakota’s governor turned to an old friend
who had built an app for a college football team in 2013.

Apple <https://quotes.wsj.com/AAPL> Inc. and Alphabet
<https://quotes.wsj.com/GOOG> Inc.’s Google deployed technology that at
least five U.S. states agreed to adopt, but integrating it into
smartphone apps takes time and comes with significant trade-offs. Some
local health departments aren’t keen on privacy restrictions in the
Apple-Google protocol
<https://www.wsj.com/articles/why-google-and-apple-stores-had-a-covid-19-app-with-ads-11591365499>
that limit information they can collect. Others had already sunk money
into Covid apps before the tech giants arrived on the scene.

Around 30 countries, including Singapore, Israel and China, have rolled
out or are planning Covid-tracking apps
<https://www.wsj.com/articles/coronavirus-contact-tracing-apps-launch-across-europe-amid-hopes-for-broad-adoption-11592319612>
at the national level. The kaleidoscope of U.S. apps, deployed at the
state or county level, risks confusing users and could make it harder to
track potential Covid-19 exposure when people travel across state lines.

“We don’t have a clear national strategy for disease control, for
traditional contact tracing, so there’s nowhere to plug in a digital
program,” said Josh Sharfstein, a Johns Hopkins professor who was part
of a team that researched contact-tracing apps. “You really need a
fundamental strategy, and then you can decide if tech will help or not.”

The Centers for Disease Control and Prevention issued high-level
guidelines on digital tracing
<https://www.cdc.gov/coronavirus/2019-ncov/downloads/php/prelim-eval-criteria-digital-contact-tracing.pdf>
in mid-May, recommending in part that apps be open source and allow
users to delete their data. It left the specifics to local governments.

A CDC spokesman said local officials “are best-positioned to know what
tools will work best for their jurisdictions.”

Even when the technology is sorted out, there are scores of questions
for local officials to consider, including: What type of information
should be collected and who should have access to it? And also, is the
data collected even useful in slowing the spread of Covid-19?

States are recruiting armies of human contact tracers who ask people who
test positive where they have been in order to track down and quarantine
others who may be infected. The painstaking process suffers when
infected people don’t share information.

There are primarily two types of apps in development to supplement human
efforts. One tracks a person’s location. The other tracks his or her
proximity to others. In the early going many states and localities
turned to the first kind of app, which uses GPS technology. The GPS apps
drop pins on a map to track where users have been, so if they test
positive they can give human contact tracers specific data about where
they might have exposed other people.

The other apps, which most states plan to build based on the smartphone
technology from Apple and Google, use Bluetooth signals to track how far
apart devices are from each other. Using the same technology that
transmits audio to wireless headphones, these apps will log when
smartphones come near each other and for how long, so that if people
test positive, others with the app who were in their vicinity in recent
days—even if an anonymous crowd on a subway, for example—can
automatically be notified. These apps don’t tell users the specific
location, time or date of their potential exposure, in order to protect
the identity of the person who potentially exposed them.

Covid-tracking apps won’t succeed unless large numbers of people install
them and agree to share their data voluntarily, officials and executives
involved in the efforts say. The first two apps released in the U.S., in
Utah and North Dakota, haven’t been widely adopted.

Dan Carvajal, a 32-year-old from Salt Lake City, downloaded Utah’s
GPS-based app, Healthy Together, and was optimistic about its potential,
but over time he realized no one he knew was using it.

“This is just a slick looking Big Brother app that isn’t installed
enough to be useful, so why am I giving them all my information when no
one else is?” he said. He deleted the app.

Twenty Labs LLC, the developer of Healthy Together, said 75,000 users,
or less than 3% of the state’s population, have installed the app. The
app has referred over 12,000 Utah residents for testing and sent them
their test results, Twenty said.

In April, two weeks after Utah signed a contract with Twenty to build
Healthy Together and integrate it with the state’s public-health
systems, George McEwan, Utah’s director of information technology,
called the app “a joke” in an email to colleagues that was obtained by
The Wall Street Journal as part of a public-records request. “This
project is all vaporware. They don’t actually have anything otherwise we
would have seen it already,” he wrote.

In response to the comments in the email, Twenty referred to a statement
from the Utah Department of Health that said Twenty “has been a strong
and reliable partner in our efforts to harness technology in an
unprecedented public-health effort.”

In late May, Twenty first delivered the software portal that was
supposed to be done in 10 days, according to Twenty and the state. After
the state tested the system and patched the integration, contact-tracers
began training to use the portal on June 15, said a spokesman for the
Utah Department of Health.

Mr. McEwan didn’t comment. A spokeswoman for the state’s information
technology department said his comments were made prior to the actual
launch of the app, and since that time the app has been “immensely
valuable for many residents in Utah.”

Joyce Schroeder, a professor at University of Arizona who is leading its
efforts in coordination with her state to roll out a digital
contact-tracing app using the Apple-Google technology, said while the
app can help contain the virus, using it shouldn’t be mandatory on campus.

“Everyone needs to be able to make their own choice about how they
function in society,” she said. “If you’re letting your phone be tracked
over time, that should be voluntary.”

Many app developers are struggling with the most fundamental element:
tracking location accurately. GPS technology, based on pings from a
network of satellites in space, is generally accurate to within about 16
feet, though depending on topography and other variables it can
sometimes be off by hundreds of feet. In urban environments, it can show
people to be in the same spot on a map if they are a dozen floors apart
in an apartment building.

Janie Schonauer, a 62-year-old resident of Garrison, N.D., downloaded
the state’s GPS-based Care19 app after Gov. Doug Burgum promoted it. “I
had a physical therapy appointment at the hospital one day. It didn’t
even mark I was there,” said Ms. Schonauer.

A few days later she drove south to Bismarck, stopping at her grandson’s
orthodontist, a McDonald’s <https://quotes.wsj.com/MCD> and a greenhouse
to pick up some pine trees, but the app didn’t log any of those stops,
she said.

Tim Brookins, a Microsoft Corp. engineer who built Care19, said user
settings, such as using battery-saver mode or turning off Wi-Fi, and
environmental conditions may have interfered with the app’s location
logging.

Care19 grew out of an app Mr. Brookins built in his spare time a few
years ago to track the locations of North Dakota State University fans
as they drove to a championship football game near Dallas. Gov. Burgum
asked him to repurpose that “Bison Tracker” app for contact-tracing
purposes. He is expected to get paid about $9,000 through September.

“I didn’t really know anything about contact tracing,” Mr. Brookins
said. He has been racing to improve its quality after negative early
reviews and fending off accusations that he violated users’ privacy by
sending their data to Foursquare Labs Inc., a location-data company, a
violation of Care19’s privacy policy. Mr. Brookins said he added
Foursquare’s code to help determine locations visited by users.
Foursquare and Mr. Brookins said the company doesn’t collect Care19
users’ data or use it for any other reason.

The GPS-based effort by a team from MIT looked promising in March, when
the professor leading it said his team was in negotiations to partner
with the World Health Organization. It formed a nonprofit, Path Check
Foundation, that pitched its app, Safe Paths, to U.S. public-health
authorities and foreign governments, with help from consulting firm
Ernst & Young LLP, according to a planning document reviewed by the Journal.

The WHO partnership hasn’t materialized. While MIT had agreements with
over a dozen health authorities as of early May, only Haiti’s
public-health authority is currently listed in the app.

A Wall Street Journal reporter who tested the app from California for a
month found that it repeatedly logged one location he had never traveled
to and failed to log others where he had. At one point it incorrectly
recorded his location hundreds of miles off the coast of Ghana in the
Gulf of Guinea.

A New York-based reporter found the app recorded her route accurately,
though it logged many more stops than she actually made.

A Massachusetts-based editor and reporter found that the app correctly
recorded local visits to his grocery store and his sister’s house but
incorrectly said he drove to a local pond. It also marked accurately
that he took a drive to Cape Cod but didn’t log that he stayed in one
location there for two hours.

The MIT app has struggled with other functionality including encrypting
data and importing location history from Google Maps. New leadership and
advisers were brought in to stabilize the project amid concerns that the
app wasn’t being developed quickly enough, according to a person
familiar with its operations.

Teton County is among several U.S. spots that have signed up for MIT’s
app. Jodie Pond, director of the Teton County public-health department,
said if Safe Paths works it could help save restaurants and other
businesses who can’t survive without the crowds of tourists who come
during summer months. “It’s not there yet—we recognize that,” Ms. Pond said.

The limitations of GPS and concerns over privacy helped propel the
Apple-Google effort to create a Bluetooth system that would anonymously
track exposures by logging phones coming near each other. Apple made
technical changes that would allow special Covid exposure-notification
apps to use Bluetooth, even when screens are dark, to register nearby
iPhones or devices using Google’s Android operating system.

Arizona plans to supplement human tracers with Covid Watch, an app using
the Apple-Google standard, developed by a team at Stanford University.

Testing in Arizona began last week, and Ms. Schroeder, the professor
leading the project, said there is still plenty of work ahead for dozens
of people on the team.

Alabama and South Carolina are among other states that signed up to use
the Apple-Google technology.

To protect users’ privacy, Apple and Google are prohibiting GPS apps
that track users’ locations from using their Bluetooth system, a blow to
developers who hoped to combine the technologies to build apps that
register both proximity and location.

Bluetooth can tell you if you were near an infected person, but only GPS
can tell you where that interaction happened—say, a small coffee shop or
an outdoor soccer field—to jog your memory and help you gauge your risk
level.

In Utah, the state’s GPS-based Healthy Together app currently logs the
locations where an opted-in user has been for at least five minutes,
along with other personal data such as the user’s name and phone number.
The app can share de-identified data with local, state and federal
officials, as well as researchers for Covid-19 response.

But in order to work with Apple and Google’s protocol, the app’s
developer, Twenty, offered to quit collecting location data—effectively
dropping its key functionality—according to correspondence written by a
Google representative to state employees that was obtained by the
Journal through a public-records request.

Twenty said it didn’t make that offer. Co-founder and chief strategy
officer Jared Allgood said the matter hadn’t been decided.

An Apple spokesman declined to comment on the Utah app specifically, but
said no exceptions have been made for location-gathering apps. Google
declined to comment.

As a result of the Apple-Google restrictions, local officials in some
states said they might need to deploy two separate apps.

Brendan Babb, Anchorage, Alaska’s chief innovation officer, signed a
letter of intent to work with the MIT team for its GPS app but is also
looking into Covid Watch, the Bluetooth app that is being prepared for
use in Arizona.

It’s hard enough for people to download one app, let alone two, said Mr.
Babb. “It’s a big ask for my residents,” he said.

North Dakota is having Mr. Brookins, the developer of its Care19 app,
build a second app based on Google and Apple’s Bluetooth protocol.

Washington state’s digital tracking program, which is changing its name
to CommonCircle, also plans to release two different apps. So does the
MIT team.

Some officials aren’t racing to adopt smartphone apps. New York is
focusing on building a statewide force to do traditional
person-to-person contact tracing first.

After that it will examine which smartphone app might be right for the
state, said Larry Schwartz, a former chief of staff to Gov. Andrew Cuomo
who was recruited to help with the initiative.


-- 

Regards
brd

Bernard Robertson-Dunn
Canberra Australia
email: brd at iimetro.com.au