Estimating Public Bicycle Trip Characteristics with Consideration of Built Environment Data

A reliable estimation of public bicycle trip characteristics, especially trip distribution and duration, can help decision-makers plan for the relevant transport infrastructures and assist operators in addressing issues related to bicycle imbalance. Past research studies have attempted to understand the relationship between public bicycle trip generation, trip attraction and factors such as built environment, weather, population density, etc. However, these studies typically did not include trip distribution, duration, and detailed information on the built environment. This paper aims to estimate public bicycle daily trip characteristics, i.e., trip generation, trip attraction, trip distribution, and duration using points of interest and smart card data from Nanjing, China. Negative binomial regression models were developed to examine the effect of built environment on public bicycle usage. Totally fifteen types of points of interest (POIs) data are investigated and factors such as residence, employment, entertainment, and metro station are found to be statistically significant. The results showed that 300 m buffer POIs of residence, employment, entertainment, restaurant, bus stop, metro station, amenity, and school have significantly positive effects on public bicycle generation and attraction, while, counterintuitively, 300 m buffer POIs of shopping, parks, attractions, sports, and hospital have significantly negative effects. Specifically, an increase of 1% in the trip distance leads to a 2.36% decrease in the origin-destination (OD) trips or a 0.54% increase of the trip duration. We also found that a 1% increase in the number of other nearby stations can help reduce 0.19% of the OD trips. The results from this paper can offer useful insights to operators in better estimating public bicycle usage and providing reliable services that can improve ridership.

Automated summary

This study estimated daily trip characteristics of public bicycles using points of interest and smart card data from Nanjing, China, including trip generation, attraction, distribution, and duration. The results showed that factors in the built environment have significant effects on public bicycle usage, with certain points of interest like residence, employment, entertainment, and metro stations having positive impacts while others like shopping and parks having negative impacts. Increased trip distance led to decreased origin-destination trips and increased trip duration, while more nearby stations were associated with a reduction in OD trips.