Understanding the spatiotemporal dynamics of ride-hailing services: A study of demand and supply patterns using a large-scale driver activity dataset

Abstract

As ride-hailing services have significantly changed the transportation landscape, understanding their operations becomes crucial for efficient urban planning and policymaking. Despite the growing number of studies on ride-hailing services, there is a significant research gap in exploring supply-side characteristics at granular spatiotemporal scales due to data limitations. This paper comprehensively analyzes spatiotemporal patterns of ride-hailing services from the demand, supply, and interactions perspectives. The study uses a comprehensive dataset comprising both driver activity data and trip order data from all ride-hailing platforms in the City of Toronto. Several key system performance indicators are examined using the large-scale platform dataset, including passenger wait time, driver wait time, trip confirmation time, pickup delay, demand–supply ratio, idle distance ratio, and active trip time ratio. Statistical analyses including t-tests and Moran’s I index have been used to quantify temporal and spatial variations. The analysis reveals significant temporal and spatial heterogeneity in ride-hailing characteristics, which suggests the need for targeted policies and planning for effective urban transportation. In addition, this study conducts a Pearson correlation analysis to quantify the correlation between ride-hailing performance and aggregated socioeconomic characteristics. These insights can inform efficient fleet management strategies, facilitate dynamic pricing decisions, and enable ride-hailing companies to enhance service quality.