Bridge health monitoring using existing telecommunication fiber-optic networks

Abstract Aging bridges pose significant risks to public safety. Although existing sensing technologies offer a potential solution for structural health monitoring, the high cost and complexity of installing and maintaining sensing systems have limited their scalability and long-term deployment. Here, we present a scalable and cost-effective approach that turns pre-existing telecommunication (telecom) fiber-optic cables into a dense array of dynamic strain sensors to continuously monitor bridges. Using telecom fibers, we retrieve guided wavefields across bridge superstructures, which was impossible with existing sensing technologies due to their limited scalability and the prohibitive cost of ultra-dense sensor arrays required to capture meter-scale wavelengths. We also estimate quasi-static displacements with millimeter-level accuracy and extract modal parameters. These results demonstrate that our approach is robust to noise from imperfect fiber-bridge coupling, a limitation inherent to the non-dedicated nature of telecom fibers used as sensors. Furthermore, we demonstrated the effectiveness of this approach through field evaluations across six bridges in the United States and South Korea. Because telecom fibers are widely deployed and co-located with civil infrastructure systems, this study introduces a new sensing paradigm that leverages existing fiber-optic networks to enable large-scale, real-time structural monitoring without the cost and logistical burden of dedicated sensors.