Integrating Pre-Existing Telecommunication Fiber Cable Vibration Sensing and Drive-by Vehicle Vibration Sensing for Scalable Bridge Health Monitoring

Abstract

We introduce a novel and scalable bridge health monitoring approach by combining pre-existing telecommunication fiber vibration sensing and drive-by vehicle vibration sensing. Many bridges have pre-existing telecommunication fiber cables crossing in conduits, which can function as a distributed acoustic sensing system to continuously record bridge dynamic strain responses in high spatiotemporal resolution. In addition, the drive-by vehicle vibrations capture the information of the input load to the bridge and the interactive bridge vibrations with high spatial resolution. Our approach integrates the two sensing systems to offer complementary input-output information and insights into bridge conditions. It estimates bridge modal properties by modeling the vehicle-bridgefiber interaction with an augmented state-space model. The model considers the bridge dynamics and input loads as a joint state observed by the telecommunication fiber vibration and drive-by vehicle vibration sensing systems, respectively. Our approach is scalable as it does not require on-site installation and maintenance of sensing systems on the bridge. It is also more accurate than using a single sensing system because it takes into account both the input and output signals of the system being analyzed. We evaluate our approach with real-world experiments on a steel stringer/multi-beam bridge in Palo Alto, California, with telecommunication fiber cables running under the deck. Our approach successfully estimates bridge dynamic properties, including natural frequency, damping ratio, and mode shape. It outperforms a baseline method that only uses telecommunication fiber vibrations, which indicates the potential for accurate and scalable bridge health monitoring.