A Spatiotemporal Analysis of Shared Micromobility Trips in First- and Last-Mile Public Transit Integration

Abstract

Shared micromobility services, such as electric scooters and electric bikes, have the potential to address first- and last-mile challenges of transit networks, and planners need to understand the factors driving the demand for these services. This study addresses that need through exploratory and regression analysis. First, we examined various catchment area sizes around transit stations to identify integrated trips. Subsequently, we conducted descriptive analysis and k-means clustering to investigate the spatiotemporal pattern of first- and last-mile trips. In the final phase, we developed regression models to identify factors influencing demand for these trips, using datasets collected from Calgary, Canada. The analysis revealed that during weekday morning hours, first-mile trips outnumbered last-mile trips, while this trend reversed during the afternoon hours. On average, the integration ratio was about 20.6% for first-mile trips and 21.9% for last-mile trips. The cluster analysis uncovered temporal and spatial variations: central urban areas experience peak last-mile trips during morning hours, while residential zones show higher rates of first-mile trips. The regression results highlighted the factors influencing first- and last-mile demand, unveiling temporal patterns. For instance, private dwelling density was associated with higher first-mile demand in the morning and last-mile demand in the afternoon. Additionally, points of interest (POI) positively correlated with first-mile demand in the evening and last-mile demand in the morning. Bike infrastructures and path connectors were also strong positive predictors across time periods. The results offer insights into multimodal transport planning, emphasizing the importance of considering demand variation and optimizing operations to support vehicle rebalancing.