Bicycle trips in Endomondo, Google Maps, and MapQuest: A comparison between South Florida and North Holland

ABSTRACT

Comparison of routes recommended by online trip planners, such as Google Maps and MapQuest, with routes traveled by cyclists allows estimation of whether the routing criteria used in such trip planners are the same as those used by cyclists. This study uses GPS tracking data from the Endomondo fitness tracker app from trips in Miami-Dade (Florida) and North Holland and compares their characteristics to cycling routes suggested by Google Maps, MapQuest and the shortest path. Results highlight which trip attributes differ between Endomondo and its alternative trips, and how consistent those differences are across the two study areas. For example, Endomondo cyclists in both regions go through fewer traffic signals, and use more cycleways, footways, or bike lanes than suggested routes from Google and MapQuest. A multinomial logit model on observed Endomondo trips provides insight into the routing preferences of Endomondo commute and sport cyclists in both study regions.

KEYWORDS:

• Fitness tracker app
• Endomondo
• route choice
• cycling
• trip planner

Acknowledgments

The corresponding author would like to acknowledge the Austrian Marshall Plan Foundation for providing study-abroad scholarship money to conduct this research.

Data availability

Raw data available at https://www.github.com/mathwiz961/routing

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. https://github.com/joh12041/route-externalities/tree/master/mapping_platforms

2. https://developer.mapquest.com/documentation/directions-api/

3. https://gis-mdc.opendata.arcgis.com/

4. https://gis-fdot.opendata.arcgis.com/

5. https://land.copernicus.eu/pan-european/corine-land-cover