top of page

New Publication on Urban Traffic Monitoring Using Distributed Acoustic Sensing and Deep Learning in Journal of Transport Geography

Nov 13
2 min read

CAI Lab is pleased to announce a new publication led by Ph.D. Candidate Hao Tian. The article, “Monitoring Urban Traffic Dynamics at Fine Spatiotemporal Resolution Using Distributed Acoustic Sensing and Deep Learning,” has been published in Journal of Transport Geography.


This research explores the potential of distributed acoustic sensing (DAS) as an innovative approach to urban traffic monitoring. By transforming underground fiber-optic cables into continuous sensing arrays, DAS captures ground vibrations generated by passing vehicles, offering a passive and privacy-preserving approach to monitoring traffic activity at fine spatial and temporal scales. Using a fiber-optic sensing network in College Station, Texas, the study integrates DAS with deep learning models to detect vehicle trajectories, estimate traffic conditions, and examine how urban traffic patterns respond to major events.


Key Findings:

  • DAS enables high-resolution urban traffic monitoring. Underground fiber-optic cables can capture vehicle-induced vibrations at meter-level spatial and second-level temporal resolution, providing detailed observations of traffic activity along roadways.

  • Deep learning effectively extracts vehicle trajectories from DAS signals. By converting raw vibration measurements into spatiotemporal representations and applying deep learning-based object detection models, the proposed framework achieves reliable vehicle detection and traffic volume estimation, demonstrating strong agreement with manually annotated observations.

  • DAS reveals distinct traffic responses to major events. Traffic activity increased by approximately 34% during football gamedays and decreased by approximately 50% during the Thanksgiving holiday, highlighting the ability of DAS to capture event-driven changes in urban mobility.


This research demonstrates the potential of integrating distributed acoustic sensing and deep learning to transform existing fiber-optic infrastructure into a continuous urban traffic observatory. The proposed framework offers new opportunities for monitoring traffic dynamics, understanding human activities, and supporting intelligent transportation management through high-resolution, privacy-preserving sensing technologies.


For More:

Tian, H., Cai, H., Chen, X., & Yang, Y. (2026). Monitoring urban traffic dynamics at fine spatiotemporal resolution using distributed acoustic sensing and deep learning. Journal of Transport Geography. https://doi.org/10.1016/j.jtrangeo.2026.104901



 
 

City Analytics & Informatics (CAI) Lab 

Department of Geography

Texas A&M University

© 2026 City Analytics & Informatics Lab

Email

  • Twitter
  • LinkedIn

Address

3147 TAMU

College Station, TX 77843

bottom of page