Gulliver | Dodgy neighbourhoods

How to find safe streets

Researchers have developed an algorithm that attempts to predict which streets people will perceive as safe

By N.B. | Washington, DC

RESEARCHERS at the Massachusetts Institute of Technology have developed an algorithm that uses visual cues in photos to predict how safe people perceive different streets to be. It is easy to see how a site that presented this information could be useful to business travellers. It can be unsettling to find that you have wandered into a dodgy neighbourhood in an unfamiliar city. But is that what's really happening here? Streetsbloghas more:

The MIT team says their algorithm is a reliable mimic of how humans perceive visual cues in urban environments. Using a 1 to 10 scale, 84 percent of the time it can successfully predict whether real people will rate a street on the low end (less than 4.5) or the high end (more than 5.5). The factors incorporated by the algorithm are not public at this time.

The perception of “safety” that Streetscore approximates is defined vaguely, since the survey doesn’t explicitly distinguish between traffic violence and violent crime.

The interest, of course, lies in that bolded phrase. The MIT algorithm does indeed do well at predicting how people will perceive the safety of a given street. But are those perceptions accurate? As it turns out, they are. In their paper, the researchers establish a connection between perceived safety and homicides. "People are not stupid," says Cesar Hidalgo, one of the authors. "They are very good at telling places apart" on all sorts of measures. And that's not all: as Mr Hidalgo notes, even the inaccurate perception that somewhere is unsafe can affect behaviour—causing a business traveller to avoid the area, for example.

In the long run, he says, the team is interested in using its data to generate a set of "best practices" for how to make an urban environment look safe and lively, with recommendations for specific design interventions that can improve the perception of an area. And since a large body of research suggests that the appearance of an area can affect how people behave—the famous "broken windows" theory—it could have tangible results.

Discover more

How much will Hong Kong's protests damage visitor numbers?

Tourism is a surprisingly resilient industry—but only if governments want it to be

Why Hong Kong’s airport was a good target for protesters

The streets of 19th-century Paris and the postmodern architecture of Hong Kong’s main terminal have much in common


Why trains are not always as green as they seem

The “flight-shame” movement encourages travellers to go by train instead of plane. But not all rail lines are environmentally friendly