Digging Deeper Into Apple's COVID-19 Mobility Trends Data

Earlier this week, Apple released a trove of mobility data, published to help cities and the public understand the real-world impacts of the COVID-19 disease on human movement across the world. This data comes a bit after Google’s publishing of Community Mobility Reports, which also seek to provide insight into behavior.

While Google’s data, due to their widespread tracking of Android device locations, is able drill into detailed points of interest such as grocery stores, transit stations, and workplaces, Apple’s data, which is based upon Apple Maps direction requests, shows a more narrow slice of user behavior. However, given the tremendous install base of iPhones globally, it’s definitely a great signal into what folks are up to, and better yet, it is accessible as a CSV, while Google’s data is locked up inside PDFs.

After getting my hands on Apple’s data, which spans 63 countries and 89 specific cities, I felt a familiar pull to build my own visualizations to better understand exactly what is happening across the globe. While I can’t say they are as beautiful as the charts on Apple’s site, they do provide a bit more depth and context. Take a look below!

Impact on Countries

Walking Heat Map - Apple COVID Data

Walking is the most universal of human activities - I figured that seeing how it has changed over time, and being able to visualize trends by country, was absolutely the first stop in visualizing this data. To avoid a nasty “spaghetti” line chart with a million series, but still convey the timeseries nature of this data, I chose to use a trended heatmap, with color dependent on a daily “Walk Index” relative to January 13th.

Some interesting takeaways:

  • Hong Kong and Macao clearly isolate before any other country, and the darkness of Macao’s line shows how strictly quarantines were adhered to
  • Japan and Taiwan have seen the least drop-off in walking activity across the world for very different reasons:
    • Taiwan restricted external travel almost immediately and aggressively traced and quarantined sick individuals while leaving the rest of the country largely open for business
    • Japan has taken longer to ban travel and eventually postpone the Olympic games, and now faces an increasing threat just as Taiwan looks to have quelled the tide of new cases

Mode Trends - Apple COVID Data

Drop-offs by Mode of Transport

Mode Shift Distributions - Apple COVID Data

Mapping Current Activity Levels

The maps below are snapshots of global activity levels by transportation mode as of April 15th, 2020. In addition to reinforcing the trends we see above, where transit usage has dropped more sharply than walking and driving, these maps point out an interesting difference between continental Europe and Scandinavia, which has stayed much more active throughout the onset of the pandemic.

Walking Activity - Apple COVID Data

Driving Activity - Apple COVID Data

Transit Activity - Apple COVID Data