Using Big Data to Design Bike Shares

Big Data  |  Transportation  |  Mobile

Bike shares are riding a wave of popularity in the intermodal transit planning community. They are great because they help address several hot-button urban challenges simultaneously; for example, congestion, air pollution, “last mile” transit gaps, and even sedentary lifestyles. From 2004 to 2015, the number of bike share systems worldwide grew from 14 to 855 – that’s over 6,000% growth!

However, there are still only 54 bike share systems1 across America’s 486 urbanized areas.2 That means there is plenty of room for expansion in the U.S. We wanted to find out how Big Data could play a role in this expansion.

 

Our Case Study: Atlanta, Georgia

 

Relay_Bike_Share.jpg

Photo by Michael Kahn, Curbed Atlanta

Our team was excited to learn that Atlanta’s Relay bike share program would expand from 100 to 500 bikes by the end of 2016, so we decided to focus on Atlanta for this analysis. The question guiding our analyis was, “Can Big Data help urban planners understand where these 500 new bikes should be located, and where regional expansions should be planned in the future?”

 

Using Big Data to Site Bike Share Stations

First, we broke the city of Atlanta into a 1km by 1km grid of analysis zones. Then, we used StreetLight InSight® to scan for vehicle trips less than 2 miles in length that started or ended in each zone. These short trips are typically good candidates for displacing vehicles with bikes.

The Atlanta heat map below (Figure 1) describes the relative volume of short trips for each zone. The darker the red, the more short trips that started or ended in the zone. It shows planners the best candidate neighborhoods for additional bike share stations: a higher volume of short vehicle trips means a higher volume of trips that can convert to bikes – at least, once there is a convenient way to bike instead of drive. This Metric can help planners prioritize where to focus final planning and modeling efforts, allowing them to spend more time planning and less time collecting data, and increasing the odds of high utilization once the stations are deployed.

 Figure 1: StreetLight InSight Heat Map of 2 Mile Trips in the City of Atlanta

atlanta_2mi_trip_start_or_end_zoom_in.jpeg

 

Using Big Data to Identify Regional Expansion Opportunities

To find other municipalities that would benefit from a regional expansion of Atlanta’s bike share system, we expanded our 1km by 1km grid of analysis zones to the broader Atlanta metro region. The heat map below (Figure 2) shows the concentration of vehicle trips less than 2 miles long that start or end in each zone.

Figure 2: StreetLight InSight Heat Map of 2 Mile Trips in the Atlanta Region

atlanta_2mi_trip_start_or_end.jpeg

 

As you can see, the municipality of Dunwoody has a high concentration of short trips, so it’s one of the more attractive options for the Atlanta bike share’s regional expansion.  

So, there you have it - two ways to use Big Data to plan and expand bike share systems.

Do you have a bike share system in your town that you wish would expand, or an idea for using Big Data in bike share planning? Let us know in the comments!

 

Sources

  1. Bikeshares: A Review of Recent Literature
  2.  U.S. 2010 Census Press Release
  3. INRIX, a StreetLight Data partner, provided the GPS data for this StreetLight InSight analysis. Click here to read more about our GPS data sources.