Understanding congested travel in urban areas

Çolak, Serdar; Lima, Antonio; González, Marta C. · 2016 · OpenAlex-citations

DOI: 10.1038/ncomms10793

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 addresses the challenge of understanding and mitigating urban traffic congestion by analyzing the interplay between road infrastructure supply and travel demand. Motivated by the limitations of traditional, costly survey-based models and the lack of systematic studies on the limits of congestion alleviation through route choice modification, the authors aim to quantify how efficiently people move across diverse cities. The research specifically investigates whether a unified metric can explain congestion levels and assesses the potential benefits of shifting from selfish routing behaviors to socially aware routing strategies. The methodology couples road network data from OpenStreetMap with travel demand profiles derived from billions of mobile phone call detail records (CDRs) for five diverse metropolitan areas: Boston and the San Francisco Bay Area (USA), Rio de Janeiro (Brazil), and Lisbon and Porto (Portugal). The authors parsed road networks to determine supply capacity and mined mobile data to detect home and work locations, infer trip tables, and estimate morning peak-hour vehicular volumes. They compared inferred travel times with online map provider estimates and modeled route choices using game-theoretic frameworks, specifically comparing user equilibrium (selfish routing) against socially optimal flows. A generalized selfish routing model was implemented to simulate varying levels of driver awareness regarding social good. The primary finding is that a dimensionless ratio, $G$, representing the ratio of vehicular travel demand to road infrastructure supply, effectively explains the percentage of time lost to congestion across all studied cities. While population density and its spatial distribution also influence congestion, $G$ serves as the main determinant. The study found that commuting distances follow a lognormal distribution, and effective travel speeds decay significantly under congestion in high-demand cities like Rio and the Bay Area, whereas lower-density cities like Porto and Lisbon experience less speed reduction. Furthermore, the analysis of routing behaviors revealed that moderate levels of social awareness are sufficient to achieve significant collective savings. Specifically, when drivers value social good at just 10–20% of their personal cost, the system realizes 40–60% of the potential travel time savings compared to purely selfish routing. The socially optimal state is approached when the weight of social good reaches 0.5. The significance of this work lies in providing a scalable, data-driven framework for understanding urban mobility that transcends individual city characteristics. By establishing $G$ as a key predictor of congestion, the study offers policymakers a clear metric for assessing infrastructure adequacy. Additionally, the findings demonstrate that leveraging information technologies to guide drivers toward socially aware routing choices can yield substantial efficiency gains without requiring full compliance or extreme behavioral changes. This suggests that moderate interventions in route guidance systems could significantly reduce collective travel times and improve urban traffic flow.

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.

• traffic density