Spatial analysis of shared e-scooter trips

Abstract

Shared e-scooters have become a common mode of transportation in many cities around the world. E-sooters provide convenient and quick rides for short distances and can act as a connection for first/last mile trips. To date, limited studies have explored the spatial variation of e-scooter trips and there is knowledge to be gained by investigating variables associated with e-scooter trip generation. This study implemented a spatial analysis approach, Geographical Weighted Regression (GWR), to explore how factors relating to demographics, density, diversity, design, urbanism scores, distance to transit and other transportation-related variables influence e-scooter trips in Louisville, KY. More than 400,000 e-scooter trips across 159 Traffic Analysis Zones (TAZs) were included in the study. Results show TAZ-level factors including land use, age distribution, gender distribution, Walk Score and Park Score impacted the density of e-scooters trips in the TAZ. The GWR model showed improvements over a global Ordinary Least Squares (OLS) model. Local goodness of fit ranged from 0.732–0.895 across the study area. This study can help governments and e-scooter sharing companies develop policies that maximize e-scooter use, equity, and accessibility while improving the mobility of cities.