Data-Driven Detection of Anomalies and Cascading Failures in Traffic Networks

Sanchita Basak, Afiya Aman, Aron Laszka, Abhishek Dubey, Bruno Leao

Proceedings of the 11th Annual Conference of the Prognostics and Health Management Society (PHM) 2019

PDF DOI

Why This Matters

Understanding cascading failures in traffic networks is critical for mitigating widespread disruptions and improving system resilience. This work is innovative because it combines multiple modeling approaches (physical models and data-driven learning) with specialized techniques for capturing spatial-temporal dependencies in traffic data. The Timed Failure Propagation Graph provides a novel mechanism to identify congestion sources and understand how local incidents affect larger network regions.

What We Did

This paper develops methods for detecting anomalies and cascading failures in traffic networks using a combination of model-driven and data-driven approaches. The work builds LSTM-based traffic speed predictors for each road segment while using Gaussian Process Regression as a comparison baseline. A Timed Failure Propagation Graph is formulated to identify root causes of congestion and trace how failures cascade through the network.

Key Results

The LSTM-based speed predictors achieved better performance than Gaussian Process Regression with average precision of 0.8507 on the precision-recall curve. The system successfully identified cascading effects of traffic congestion with average precision of 0.9269 and recall of 0.9118 when tested on ten congestion events in Nashville. The approach demonstrated the ability to isolate root causes of network-wide congestion from complex spatial-temporal data.

Full Abstract

Traffic networks are one of the most critical infrastructures for any community. The increasing integration of smart and connected sensors in traffic networks provides researchers with unique opportunities to study the dynamics of this critical community infrastructure. Our focus in this paper is on the failure dynamics of traffic networks. By failure, we mean in this domain the hindrance of the normal operation of a traffic network due to cyber anomalies or physical incidents that cause cascaded congestion throughout the network. We are specifically interested in analyzing the cascade effects of traffic congestion caused by physical incidents, focusing on developing mechanisms to isolate and identify the source of a congestion. To analyze failure propagation, it is crucial to develop (a) monitors that can identify an anomaly and (b) a model to capture the dynamics of anomaly propagation. In this paper, we use real traffic data from Nashville, TN to demonstrate a novel anomaly detector and a Timed Failure Propagation Graph based diagnostics mechanism. Our novelty lies in the ability to capture the the spatial information and the interconnections of the traffic network as well as the use of recurrent neural network architectures to learn and predict the operation of a graph edge as a function of its immediate peers, including both incoming and outgoing branches. Our results show that our LSTM-based traffic-speed predictors attain an average mean squared error of $6.55\times10^-4$ on predicting normalized traffic speed, while Gaussian Process Regression based predictors attain a much higher average mean squared error of $1.78\times10^-2$. We are also able to detect anomalies with high precision and recall, resulting in an AUC (Area Under Curve) of 0.8507 for the precision-recall curve. To study physical traffic incidents, we augment the real data with simulated data generated using SUMO, a traffic simulator. Finally, we analyzed the cascading effect of the congestion propagation by formulating the problem as a Timed Failure Propagation Graph, which led us in identifying the source of a failure/congestion accurately.

Cite This Paper

@inproceedings{Basak2019b,
  author = {Basak, Sanchita and Aman, Afiya and Laszka, Aron and Dubey, Abhishek and Leao, Bruno},
  booktitle = {Proceedings of the 11th Annual Conference of the Prognostics and Health Management Society (PHM)},
  title = {Data-Driven Detection of Anomalies and Cascading Failures in Traffic Networks},
  year = {2019},
  month = {oct},
  abstract = {Traffic networks are one of the most critical infrastructures for any community. The increasing integration of smart and connected sensors in traffic networks provides researchers with unique opportunities to study the dynamics of this critical community infrastructure. Our focus in this paper is on the failure dynamics of traffic networks. By failure, we mean in this domain the hindrance of the normal operation of a traffic network due to cyber anomalies or physical incidents that cause cascaded congestion throughout the network. We are specifically interested in analyzing the cascade effects of traffic congestion caused by physical incidents, focusing on developing mechanisms to isolate and identify the source of a congestion. To analyze failure propagation, it is crucial to develop (a) monitors that can identify an anomaly and (b) a model to capture the dynamics of anomaly propagation. In this paper, we use real traffic data from Nashville, TN to demonstrate a novel anomaly detector and a Timed Failure Propagation Graph based diagnostics mechanism.  Our novelty lies in the ability to capture the the spatial information and the interconnections of the traffic network as well as the use of recurrent neural network architectures to learn and predict the operation of a graph edge as a function of its immediate peers, including both incoming and outgoing branches.   Our results show that our LSTM-based traffic-speed predictors attain an average mean squared error of $6.55\times10^{-4}$ on predicting normalized traffic speed, while Gaussian Process Regression based predictors attain a much higher average mean squared error of $1.78\times10^{-2}$. We are also able to detect anomalies with high precision and recall, resulting in an AUC (Area Under Curve) of 0.8507 for the precision-recall curve. To study physical traffic incidents, we augment the real data with simulated data generated using SUMO, a traffic simulator.  Finally, we analyzed the cascading effect of the congestion propagation by formulating the problem as  a Timed Failure Propagation Graph, which led us in identifying the source of a failure/congestion accurately.},
  attachments = {https://www.isis.vanderbilt.edu/sites/default/files/PHM_traffic_cascades_paper.pdf},
  category = {conference},
  contribution = {lead},
  doi = {https://doi.org/10.36001/phmconf.2019.v11i1.861},
  file = {:Basak2019b-Data_Driven_Detection_of_Anomalies_and_Cascading_Failures_in_Traffic_Networks.pdf:PDF},
  keywords = {anomaly detection, cascading failures, traffic networks, LSTM networks, data-driven methods, congestion forecasting},
  project = {smart-transit,smart-cities,cps-reliability},
  tag = {ai4cps,transit},
  month_numeric = {10}
}

Data-Driven Detection of Anomalies and Cascading Failures in Traffic Networks

Why This Matters

What We Did

Key Results

Full Abstract

Cite This Paper

Quick Info

Keywords

Research Areas

Search Tags