COVID-19 Digital Contact Tracing: Apple and Google Work Together as MIT Tests Validity
In a rare act of cooperation, Google and Apple this month released specifications for software developers to build digital contact tracing apps for Apple and Google mobile operating systems, which jointly encompass the majority of smartphones around the world.
Digital contact tracing, which can automatically notify an individual if they've crossed paths with someone who tested positive for COVID-19, has been proposed as a way to augment manual contact tracing, which requires the painstaking work of thousands of trained workers per state to identify, track, and assist individuals exposed to the virus.
As digital contact tracing technologies advance, two questions rise to the surface: Will state health officials and individuals opt to use the technology? And, if so, how well will it work?
In theory, digital contact tracing can be used to contact large numbers of people at a fraction of the cost of manual contact tracing. Now that Apple and Google are onboard, and the CDC has issued guidance criteria for the technology, it seems poised for use. But cities and states across the United States have yet to adopt the digital tool, instead focusing on manual tracing, according to a recent WIRED investigation.
In Massachusetts, the first state to initiate a major manual contact tracing program, Governor Charlie Baker has said publicly that there are conversations about incorporating smartphone data into the manual program, but it must be done in a way that protects privacy and confidentiality. Other government leaders have voiced similar privacy concerns. Smartphone tracking programs are being effectively employed in other countries to slow the spread of the virus, but many, like TraceTogether in Singapore, would violate privacy laws if enacted in the United States.
Apple and Google's effort is certainly focused on privacy. In their joint draft documentation [PDF], Apple and Google said that apps using their contact tracing application programming interfaces (APIs) must be made by or for the use of government health authorities; users must be opt-in only and shall consent before sharing a positive test result; gathered information is only for use for COVID-19 exposure information, so cannot be used for advertising or any other purposes; and the companies have pledged to discontinue the use of the system once the crisis has passed.
The two companies are being advised in part by the MIT-led Private Automated Contact Tracing, or PACT project, one of the international research teams pioneering the Bluetooth-based privacy protocol at the heart of Apple and Google's solution. The PACT system seeks to automate contact tracing by detecting and logging proximity between phones using Bluetooth signals, or chirps," from phones within an approximate 6-foot radius and picked up for a particular duration of time.
Unlike early digital contact tracing efforts from the WHO and MIT's SafePaths which relied on GPS data, the PACT system does not collect location data. Instead, the system relies on phones sending out anonymous Bluetooth chirps"-random, rotating numbers which do not reveal from where or whom they were sent. Then, if a person tests positive for the new coronavirus, they can upload all the chirps their phone has sent out in the last two weeks to a database. If any of those chirps match ones picked up by someone else's phone, a notification will inform that person of a possible exposure.
Our job is to make sure what [Apple and Google] implement is as private as possible," says Marc Zissman, co-PI on the PACT team and associate head of the Cyber Security and Information Sciences Division at MIT Lincoln Laboratory. But we're also worried about, how well does this really work?"
Bluetooth, which transmits data via low-power radio waves, was designed as a communication waveform, not for determining the distance between objects, like a radar waveform, Zissman points out. To determine if Bluetooth technology has the ability to detect contacts that could lead to COVID-19 infection-such as at 6 feet away for 15 to 30 minutes-the PACT team is now leading an international consortium of cooperating research laboratories to collect data and conduct experiments.
Some of the data are being collected with robots at the Lincoln Laboratory's Autonomous Systems Development Facility. By moving robots equipped with smartphones around a room under various conditions, the MIT team is gathering data on the signal strength of the chirps for various distances and amounts of times. That data can be used as the basis for improving machine learning algorithms to detect distance via Bluetooth without burning through battery life on devices, says Zissman. Plus, if the CDC changes its guidance on how close people have to be for how long to become infected, the team could adapt the protocol to measure the optimal distance and time.
The team has initial data demonstrating the validity of the system, Zissman notes. We're very optimistic about the whole thing, but we still have to prove it," he says. The team is making data public as they collect it, and will continue to provide recommendations to Apple and Google moving forward.
Apple and Google have released screenshots of how a digital tracing app might look to a user, but the companies have thus far deferred making an app of their own, leaving that to industry and government groups. The PACT team continues to be involved with assisting governments looking to implement the technology.
All of our stuff is available for anybody who wants to look at it," says Zissman. Our role is to show how the technology could work and then try to answer questions for government organizations who might use it."
< Back to IEEE COVID-19 Resources