Google uses location data to show which places are complying with stay-at-home orders
Google uses location data to show which places are complying with stay-at-home orders — and which aren’t
New COVID-19 mobility reports show changes in traffic to stores, parks, transit stations, and more
By Casey Newton@CaseyNewton
Apr 3, 2020, 2:00am EDT
https://cdn.vox-cdn.com/thumbor/rz0X...ge_covid.5.jpg
Sampling of data from US report. Image: Google
Google is using location data gathered from smartphones to help public health officials understand how people’s movements have changed in response to the global COVID-19 pandemic. In a blog post early Friday morning, Google announced the release of its COVID-19 Community Mobility Reports.
The reports use data from people who have opted in to storing their location history with Google to help illustrate the degree to which people are adhering to government instructions to shelter in place and, where possible, work from home.
“As global communities respond to the COVID-19 pandemic, there has been an increasing emphasis on public health strategies, like social distancing measures, to slow the rate of transmission,” the company said in a blog post.
“In Google Maps, we use aggregated, anonymized data showing how busy certain types of places are — helping identify when a local business tends to be the most crowded. We have heard from public health officials that this same type of aggregated, anonymized data could be helpful as they make critical decisions to combat COVID-19.”
Anyone can view the reports, which cover 131 countries to start. In many locations, users can search for more regional data, examining reports for individual states, provinces, and counties. After the user selects a geographic region, Google will generate a PDF with the data it has collected.
Google said that it chose PDFs over web pages because they could be more easily downloaded and shared with workers in the field.
https://cdn.vox-cdn.com/thumbor/KGCq...arch_Spain.gif
Each report contains information about movement patterns in six categories:
- Retail and recreation, covering visits to restaurants, cafes, shopping centers, theme parks, museums, libraries, movie theaters, and similar locations.
- Grocery and pharmacy, covering supermarkets, food warehouses, farmers markets, specialty food shops, and drug stores.
- Parks, covering public beaches, marinas, dog parks, plazas, and other public spaces.
- Transit stations, covering subway stops and bus and train stations.
- Workplaces, covering offices.
- Residences, covering people’s homes.
A sample report viewed by The Verge for California, where a shelter in place order has been in effect since March 19th, showed steep declines to retail and recreation locations and transit stations, with a moderate uptick in time spent at home.
Data covers the past 48 to 72 hours, Google said, and the percentage changes reflect the difference between movement this month and late January.
The move comes as technology companies have been asked by government agencies and health officials to share more data to aid in the coronavirus response.
On Saturday, the Wall Street Journal reported that mobile advertising companies were similarly sharing anonymized, aggregated data with the Centers for Disease Control and Prevention, as well state and local governments, to help officials understand the spread of the disease and coordinate their response.
Facebook has made similar data available to academic researchers.
https://cdn.vox-cdn.com/thumbor/7h-Z...eport_zoom.png
Google executives told me its program is intended to aid public health officials who need to prioritize their response based on areas of greater need.
The reports may help a county official understand that its parks remain overcrowded despite an order to shelter in place, for example — or that its parks are properly empty, but its transit stations remain too crowded. That would allow them to consider changing or amplifying messages to their communities about the need to stay away.
At the same time, a high-level look at changing mobility patterns by itself is likely to be of limited value in managing the response to the pandemic.
Countries that have had more success in fighting COVID-19 have done so by implementing aggressive testing and contact-tracing regimes, and also by making invasive use of location data.
Taiwan, for example, is using location data to create “electronic fences” around quarantined citizens, monitoring their movements to ensure they remain at home.
Google’s data does not include personally identifiable information or show the number of visits to any particular category. And it has limits: for example, it may not be able to account for people who spend time near a location as part of permitted outdoor exercise routines.
The company considered requests from public health officials to make more data available for contact tracing — using an individual’s location to identify other people who may have been around them during the time they were infectious. But Google’s location data isn’t granular enough to determine whether someone came within 6 feet of them — the distance currently thought to bring someone within risk of transmission — and it contains enough errors to make contact tracing impractical.
Google also considered using location history data to show how crowded hospitals and other medical facilities had become. But location data can’t distinguish between healthcare workers, patients, and visitors, making the value of sharing such information questionable.
Google plans to update the data in the reports in the future, it said, but at the moment has not decided when.
Separately, Google said it would collaborate with epidemiologists working on COVID-19 to update an existing dataset of aggregated, anonymized information to forecast the path of the pandemic.
https://www.theverge.com/2020/4/3/21...vid-19-privacy
