Exploring on-demand service use in large urban areas: the case of Rome

Nuzzolo, Agostino; Comi, Antonio; Polimeni, Antonio · 2019 · DOAJ

DOI: 10.5604/01.3001.0013.5681

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This study investigates the spatial and temporal demand patterns of traditional on-demand transport services, specifically taxi usage, in large urban areas using Rome as a case study. The research is motivated by the need to improve urban sustainability, reduce congestion, and optimize public transport flexibility through better understanding of on-demand service dynamics. By analyzing taxi data, the authors aim to identify key factors influencing service demand and develop forecasting models that utilize easily accessible variables, such as census data, to predict taxi requests in specific time intervals. The methodology relies on a dataset of floating car data (FCD) collected from approximately 7,700 taxis in Rome during February 2014, comprising over 750,000 GPS records. The authors processed this data to reconstruct trip origins, destinations, travel times, and distances by identifying vehicle status (moving vs. stationary) and distinguishing trips with customers from empty runs. The analysis focused on weekdays within the city’s main road ring, dividing the area into 99 traffic zones. To model demand, the researchers employed Classification and Regression Trees (CART) to identify significant predictors and subsequently developed multiple linear regression models for six distinct time slices. Predictors included zonal population, travel attributes, metro accessibility, and employment figures across various sectors, including professional/technical activities, commerce, and hospitality. The results reveal distinct spatial and temporal patterns in taxi demand. Spatially, demand is concentrated in the city center and around major transport hubs, particularly the Roma Termini and Roma Tiburtina railway stations, which account for approximately 24% of total requests. Temporally, six specific time slices were identified, reflecting usage related to night-time metro closures, early morning commutes, and daytime tourism or business activities. The regression tree analysis identified the number of employees in professional, scientific, and technical activities as the primary predictor of demand, followed by time slice and hospitality sector employment. The resulting regression models demonstrated statistical significance, with R² values ranging from 0.20 to 0.61 across different time periods, indicating that land-use characteristics and socio-economic data are effective proxies for forecasting taxi demand. The significance of this work lies in its provision of a robust methodology for forecasting on-demand service demand using readily available urban data. The findings offer practical tools for urban planners and taxi operators to optimize service locations, improve vehicle utilization rates, and manage supply more efficiently. Furthermore, the study highlights the strong correlation between specific land-use types (particularly professional services and tourism) and taxi demand, providing insights that can inform broader urban mobility strategies and the integration of on-demand services into public transport systems.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• exposure measurement