Understanding bike-sharing usage and built environment influences across age groups: A spatial machine learning approach

Dockless bike-sharing systems (DBSs) have become an integral component of sustainable urban transportation, offering flexible, affordable, and low-carbon mobility options. Yet demographic groups use these systems differently, especially older adults whose mobility needs diverge from those of younger users. While understanding such differences is vital for inclusive transport design, little is known about spatiotemporal usage disparities or how built environments shape them. This study addresses this gap by examining both the spatiotemporal usage patterns of older adults (aged 60+) versus younger users (18–59) and the nonlinear, spatially varying impacts of built environment factors on these differences through comparative modeling. Using a dataset of over 8 million DBS trips from 700,000 unique users in Harbin, China, we develop geographically weighted boosting (GWBoost) model that integrates extreme gradient boosting (XGBoost) with geographical weighting regression (GWR) model to simultaneously capture nonlinearity and spatial heterogeneity. The results indicate that older adults' DBS trips are more geographically concentrated in Harbin's commercial and historic core and exhibit more temporally uniform patterns, compared to younger adults. Modeling results further reveal that distance to the city center and distance to metro stations consistently rank as the most influential predictors of DBS usage for both age groups, with built environment effects exhibiting pronounced nonlinear, threshold, and spatially heterogeneous patterns. It shows that limited public service provision and fine-grained street connectivity are associated with higher DBS adoption among older adults in central urban areas, whereas contrasting effects are observed in peripheral zones. As China enters an aging society, these findings underscore the importance of adapting DBS infrastructure in ways that support age-friendly and inclusive planning across diverse urban contexts.