← Back to Sessions

Monitoring Vehicle Traffic With Seismoacoustic Data Using Machine Learning

Description: 

Vehicle traffic generates detectable seismic and acoustic signals that can be recorded by nearby sensors. These recorded seismoacoustic signals allow us to monitor the movement of vehicles. Using seismoacoustic data for traffic monitoring has advantages over traditional methods, such as the use of cameras, in industrial or research facilities. Seismic and acoustic sensors have a small footprint and can be less obtrusive, as well as being less affected by light conditions than cameras. The sensors can be easily deployed in a variety of locations, making them well-suited for monitoring traffic. We demonstrate the use of machine learning for traffic monitoring using seismic and acoustic data collected from a network of sensors deployed on an industrial facility in northern Texas. Traffic-related events were detected using the short-term-average over long-term-average algorithm. These signals are characterized by short pulses (few to several seconds) with energy concentrated above 100 Hz. Because each detection corresponds to a different vehicle moving at different speeds, the similarity between detections is low. However, the detections show clear interrelated spatial and temporal patterns. Stations near entrance gates show fewer detections on weekends than on workdays. Also, the hourly distribution of events agrees well with the work schedule. We are developing machine learning algorithms such as artificial neural networks to automatically distinguish traffic-related signals from noise. The use of machine learning will allow for near real-time processing of the seismoacoustic data and quick and accurate detection of vehicles. With field data from a functioning facility, we will be able to test our algorithms with real-life scenarios.

Session: Detecting, Locating, Characterizing and Monitoring Non-earthquake Seismoacoustic Sources [Poster]

Type: Poster

Date: 4/19/2023

Presentation Time: 08:00 AM (local time)

Presenting Author: Chengping Chai

Student Presenter: No

Invited Presentation: 

Authors