Google Street View Images Could Help Us Address a Public Health Crisis

Pigeon people. Sword fights. Segway gangs.

You can see a lot of Google Street View (GSV). It's an excellent way to catch a glimpse of modern life in all its weird, wonderful glory. According to a new study, it could also be a great resource for improving public health.

Researchers analyzed thousands of GSV images to determine whether the service could provide accurate data about a population's transportation habits, according to a study published Wednesday in the journal PLOS ONE. They assert this data could help public health officials predict whether new transportation policies could make people healthier.

The World Health Organization (WHO) estimates that physical inactivity causes as many as 3.2 million premature deaths every year. Finding ways to encourage "active transport," such as walking and cycling, is one way health officials can help increase a population's activity level and potentially avoid some of those deaths.

Before public health officials can craft such policies, though, they need accurate data on existing travel patterns. This study is necessary, the researchers note, because data like this is impractical to gather, done only once every 10 years, and less detailed than what's possible from GSV. As James Woodcock, MRC Epidemiology Unit and the paper's senior author, said in a press release, these images are "freely available" and collected "in a more consistent way than many traditional surveys."

What does exist pretty much everywhere? Google Street View.

GSV includes images from more than 100 nations spanning every continent, including many low- and middle-income countries. Those are where the WHO claims 2.6 million premature deaths caused by physical inactivity occur every year.

The British team focused on 34 cities in the United Kingdom. They pulled 2,000 GSV images taken between 2010 and 2012 from a total of 1,000 random spots in each city. Then, they started counting anything they could link to transportation: pedestrians, parked bicycles, in-use bicycles, motorcycles, buses, and automobiles.

Once they knew how many people were using each method in total, they could compare the “levels” or ratio of each method to the levels documented in the U.K.'s 2011 Census and the Active People Surveys, which focus on physical activity, between 2010 and 2012. This functioned as a control.

The researchers' estimates for bicycle, public transportation, and motorbike use had a "strong correlation" to the control data. Their walking estimates, though, weren't quite as accurate, only showing "moderate" agreement. Additionally, they noted "promising results" in a pilot analysis designed to predict gender distribution amongst cyclists.

This data is already available in many parts of the world, so policy makers just have to access and analyze it.

If health officials take advantage of this wealth of data, as the researchers suggest they do, the same service that delivered this creepy shot could help save lives across the globe.