Projects urban analytics

Traffic Volume Estimation

Traffic volume estimation framework combining floating car data, traffic cameras, and network flow analysis.

Role: Co-author
Published: Discover Civil Engineering, 2026 (accepted)

Calibrated traffic volume on every road segment of the Manhattan study network through the day, beside the 8 AM probe-vehicle counts that feed the model. Re-rendered from the estimates in Fig. 9 of the paper.

This project estimates network-wide traffic volumes by combining floating car data, municipal traffic camera observations, and traffic flow modeling.

The paper, accepted for publication in Discover Civil Engineering, frames the problem as a hybrid urban sensing system: Cellular transmission model features, graph neural networks, topology-informed propagation, and ensemble square-root filtering are combined to estimate and forecast traffic volumes across a Manhattan road network.

Project details

Keywords: Floating car data

Traffic cameras

Graph neural networks

Network flow

Data assimilation

Manhattan traffic

Urban sensing

References

2026

Harnessing Floating Car Data, Traffic Camera Observations, and Network Flow Analysis for Traffic Volume Estimation

Antonina Kosikova, Mehmet Kerem Turkcan, Ahmed Darrat, and Andrew Smyth

Discover Civil Engineering, Oct 2026

Accepted for publication

arXiv DOI HTML

This work fuses floating car trajectory data, municipal traffic camera observations, cellular transmission model features, graph neural networks, and ensemble square-root filtering to estimate and forecast network-wide traffic volumes.