Communications of the ACM
Crowd-Sourced Apps Help Planners Design Better Paths for Cyclists
Crowd-sourced data can tell where bicyclists travel, a useful tool for traffic planners as they decide where to build roads and bike paths, new University of Florida research shows.
The research from the UF Institute of Food and Agricultural Sciences focused on using Strava, a GPS-based app, to analyze bicycling patterns in Miami-Dade County, Florida.
Nationally, most bicycle commutes take 10 to 14 minutes, according to 2014 Census Bureau data on bike commuting. Commuting time is at least one factor that affects how many people ride bicycles to work or other activities, researchers say. The bicycle commuter rate in Miami-Dade County is 1 percent—among the lower rates among U.S. cities. Other cities range from 0.2 percent to 21.1 percent, says Henry Hochmair, the UF/IFAS faculty member who led the new study.
Hochmair and his collaborators describe their work in "Estimating Bicycle Trip Volume for Miami-Dade County from Strava Tracking Data," published in the Journal of Transport Geography.
Using Strava, the UF/IFAS researchers found several important bicycle-riding patterns in Miami-Dade. Included among those were:
- Simply adding a bike lane to any so-called "arterial"—or major—road does not necessarily attract more cyclists. This means that more advanced methods of bicycle infrastructure improvements, such as buffered bike lanes, are needed for these types of busy roads.
- Cycling decreases the farther away from the coastline people are—starting at about 5 miles inland—for both commute and non-commute trips. For planners, this means that safe access to beaches for cyclists needs to be improved since it attracts many cyclists. It is often difficult to safely travel from inland to the beach area in the east of the county because of a lack of bicycle infrastructure in the wider east-west roads.
"The study also demonstrates that people who use mobile devices can contribute to a useful large-scale data collection of cycling patterns for urban and transportation planners, besides the fun factor that comes with logging, reviewing, and sharing one's trips," says Hochmair, a UF/IFAS associate professor of geomatics.
Similar to Google, which largely derives information about traffic jams from Android phones and hence helps car drivers optimize their trips when they use the traffic layer on Google Maps to avoid traffic jams, the cycling community hopefully can also benefit from this crowd-sourced data collection on mobile devices through better-informed decisions from policy makers and planners.
"In our digitally connected world, where large parts of the population carry a GPS-enabled mobile device, humans became living sensors who collect spatial data on a continuing basis," says Hochmair.
Data from bicycle-tracking apps, like Strava or Endomondo, capture travel patterns from a large user base and allow planners to observe how people traverse all roads of a network, and how their behavior changes over time or on certain days, Hochmair says.
Planners can use this data, for example, to analyze if and how road infrastructure improvements, such as adding a new bike lane or closing a road lane, affect bicycle use, he says. Such data can also be used in combination with stationary bicycle counters, which capture all activities, but only at a few locations over a limited time.
Traffic planners, especially those in South Florida, can use the information from the new UF/IFAS study—and data derived from Strava—to find out more about the nature of trips taken by cyclists, Hochmair says.
Other authors of the Journal of Transport Geography study are Eric Bardin and Ahmed Ahmouda.
While the information researchers gathered by Strava for the study is interesting, Hochmair cautions that Strava apps are normally used by younger men with higher incomes.
"I would expect similar results in other cities as it regards to certain types of road trips, but the data would have to be tested and verified in separate studies," he says.
No entries found