Citation

Wang W, Zhu K, Liu J, Hu J, Raganvan V, Xu J, Song X. Spat. Inf. Res. 2022; 30(1): 131-142.

Copyright

(Copyright © 2022, Korean Spatial Information Society, Publisher Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s41324-021-00418-1

PMID

unavailable

Abstract

Mapping traffic speed on road networks is crucial for urban traffic management and the development of intelligent transportation systems. Traditionally, information regarding traffic speed can be obtained from location-fixed sensors, such as loop detectors and cameras; however, these methods are limited to major road crosses. Recently, a considerable attention has been paid to utilizing vehicles with mobile phones as probes for collecting traffic information. This study proposes an open-source GIS approach to map traffic speeds in a road network. First, public service vehicles (PSVs) were identified from cellular network signaling data by measuring the similarity between cell-ID trajectories and bus routes. Then, the cell-ID trajectories of PSVs were refined into high-quality spatiotemporal trajectories, and projected onto the road network via heuristic global optimization. Finally, hourly traffic speed maps were computed by weighing the speeds of the PSVs in the road network. The approach was implemented using free and open source software for geospatial mapping stacks of toolkits (Python, TimescaleDB/PostGIS, Pandas/Pygmo2, and Matplotlib/Seaborn); this application demonstrated good results using cellular network signaling data and GPS trajectories collected in Huilongguan district, Beijing, China. Moreover, this demonstration illustrates that probe mobile monitoring is emerging as a critical technology for traffic monitoring supplements, which can help develop a comprehensive view of the roads and reduce the cost of monitoring a large area.

Language: en

Keywords

Cellular network signaling data; Longest common subsequence problem; Public service vehicles; Traffic speed mapping