X
Innovation

What 11.6 million tracked bicycle trips can tell us

We're faster when it's cold, or when we're on our way to work. We outpace cars, so long as it's rush hour--and especially if it's a Wednesday. Cyclists are a fickle bunch, according to new data.
Written by John Herrman, Contributor

When the French city of Lyon inaugurated its Vélo'v public bike project, it was widely regarded as an important experiment: Can a shared collection of thousands of bikes, rented by many more thousands of people and swapped between hundreds of docking stations, possibly work? Since 2005, the results have been rolling in. The program is a massive success, accounting for somewhere near 16,000 trips a day--nothing to scoff at in a city of few than 500,000 people.

In other words, given access to bikes, people will use them responsibly. Researchers at the École Normale Supérieure de Lyon, the Universite de Lyon, and École Polytechnique de Lyon are looking to the project to answer another question: How do they ride, exactly?

To find out, they looked to data (hat tip to the Physics ArXiv blog) collected by the Vélo'v bikes and base station between May of 2005 and December of 2007. In that time, riders in Lyon logged 11.6 million trips. While the bikes don't have GPS, they do have simple onboard computers that record which stations each bike has been dock at, how long it took them to get there from the last station, and how far the bike has been pedaled.

This might not sound like much, but a sufficiently large data set--even with just a few recorded variables--can tell us a lot. Indeed, the abstract of the researchers' report boasts that it's "the first robust characterization of urban bikers' behaviors." So, what did they find? First, a few of the more predicable bits:

  • Bikers are at their speediest on their way to work, on weekday mornings
  • The average trip time is about 15 minutes, at about eight miles per hour
  • Weekend afternoons are for slow rides--almost 40% more leisurely than peak times
  • Bikes take shorter, pedestrian-style paths than cars, generally, and take more shortcuts (That's a 13% distance reduction, if you're counting)
  • Cyclists are less hindered by traffic, for obvious reason

This kind of stuff just makes sense. But there are a few data points that stand out as a little, I don't know, strange?

  • Riding speeds are higher during rush hour, when traffic is more intense (and for cars, slower)
  • Cyclists speed up in colder months (9% in the evening hours, when the change is most dramatic)
  • Wednesday mornings have the fastest riders, almost every single time
  • Bikers are daring, and regularly "use sidewalks, drive the wrong way up one-way streets or use the bus / tramway lanes," against advice

So riders enjoy careening through traffic, sometimes against traffic, especially on icy cold winter Wednesdays. Huh!

Generally speaking, though, the data doesn't blow away any preconceptions about bikes. They're effective modes of transport in geographically small, crowded cities, and their use is (usually) dictated by clear factors like traffic or time of day. For a more detailed look, the report can be downloaded here.

The overarching conclusion, if there is one to be found, is that systems like Lyon's Vélo'v work--and they work well. Comprehensive data analysis will only make them more effective.

This post was originally published on Smartplanet.com

Editorial standards