TY - JOUR
T1 - Mining dockless bikeshare data for insights into cyclist behavior and preferences
T2 - Evidence from the Boston region
AU - Sadeghinasr, Bita
AU - Akhavan, Armin
AU - Furth, Peter G.
AU - Gehrke, Steven R.
AU - Wang, Qi
AU - Reardon, Timothy G.
N1 - Publisher Copyright:
© 2021 Elsevier Ltd
PY - 2021/11
Y1 - 2021/11
N2 - Emerging micromobility services provide not only new transportation options, but also valuable sources of data for researchers and planners seeking to understand traveler behavior and preferences and to improve transportation networks. Dockless systems for shared bicycles and electric scooters have recently begun operating in many American cities, providing data not only on trip start and end, but also on the route followed. In this study, we analyze data from more than 110,000 dockless bikeshare trips in Boston's suburbs. About 15% of trip ends were within 100 m of a transit station, indicating that bikeshare serves many more functions than access to transit. More than 40% ended in a town or village center, suggesting that bikeshare may support the local economy. From examining GPS traces, it was clear that – contrary to the assumptions of common bike routing algorithms – bike riders often use sidewalks, other footways and soft-surfaced paths, parking lots, and driveways, and 36% ride the wrong way on one-way streets. An examination of the streets with high contraflow volumes indicates two common profiles. One is streets whose travel channel is so narrow that either the bike or vehicle has to pull into a parking lane to let the other pass; those streets typically have low traffic speeds and very low traffic volumes. The other is streets whose travel channel is wide enough for a bike and car to pass without either yielding to the other; they typically have greater vehicular speed and volume, but still only one lane. After classifying streets by level of traffic stress, we find that only 7 percent of trips use exclusively low-stress links, and that about 40 percent of bike-miles are ridden on low-stress links. These low percentages are consistent with an underdeveloped bike network that forces riders to use through streets for all except ultra-local trips; it may also reflect a prevalence of sidewalk riding on major arterials.
AB - Emerging micromobility services provide not only new transportation options, but also valuable sources of data for researchers and planners seeking to understand traveler behavior and preferences and to improve transportation networks. Dockless systems for shared bicycles and electric scooters have recently begun operating in many American cities, providing data not only on trip start and end, but also on the route followed. In this study, we analyze data from more than 110,000 dockless bikeshare trips in Boston's suburbs. About 15% of trip ends were within 100 m of a transit station, indicating that bikeshare serves many more functions than access to transit. More than 40% ended in a town or village center, suggesting that bikeshare may support the local economy. From examining GPS traces, it was clear that – contrary to the assumptions of common bike routing algorithms – bike riders often use sidewalks, other footways and soft-surfaced paths, parking lots, and driveways, and 36% ride the wrong way on one-way streets. An examination of the streets with high contraflow volumes indicates two common profiles. One is streets whose travel channel is so narrow that either the bike or vehicle has to pull into a parking lane to let the other pass; those streets typically have low traffic speeds and very low traffic volumes. The other is streets whose travel channel is wide enough for a bike and car to pass without either yielding to the other; they typically have greater vehicular speed and volume, but still only one lane. After classifying streets by level of traffic stress, we find that only 7 percent of trips use exclusively low-stress links, and that about 40 percent of bike-miles are ridden on low-stress links. These low percentages are consistent with an underdeveloped bike network that forces riders to use through streets for all except ultra-local trips; it may also reflect a prevalence of sidewalk riding on major arterials.
KW - Bicycle routing, detour
KW - Bicycling
KW - Contraflow
KW - Level of traffic stress
KW - dockless bikeshare
UR - http://www.scopus.com/inward/record.url?scp=85116057917&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85116057917&partnerID=8YFLogxK
U2 - 10.1016/j.trd.2021.103044
DO - 10.1016/j.trd.2021.103044
M3 - Article
AN - SCOPUS:85116057917
SN - 1361-9209
VL - 100
JO - Transportation Research Part D: Transport and Environment
JF - Transportation Research Part D: Transport and Environment
M1 - 103044
ER -