DO VISUALLY PERCEIVED DESIGN QUALITIES INFLUENCE DOCKLESS BIKESHARE CYCLING ROUTES?

Abstract

Understanding the influences of the built environment (BE) characteristics on the dockless bike-sharing system (DBS) is crucial for supporting and developing sustainable transportation mode. Previous studies on DBS cycling have primarily investigated the effects of macro-level BE characteristics (e.g., land use) or limited street features (e.g., greenery), overlooking that of perceived street design qualities such as enclosure. To better understand whether and how street-level environment characteristics, especially perceived street design qualities, influence DBS cycling routes, we calculate cycling volume based on GPS trajectories in Ithaca, a small town in New York State, and then quantify visual features and perceived design qualities using street view imagery (SVI) and computer vision (CV). Our analysis, employing linear regression and spatial regression models while controlling macro-environmental attributes as covariates, reveal the significant association between perceived design qualities and DBS.