Recently, researchers from the National and Local Joint Engineering Center for Smart Construction and Operation of Southeast University (hereinafter referred to as the center) have been shuttling through the streets and fiber optic access machine rooms in Nanjing, using distributed fiber optic sensing technology to perform CT scans on roads and prevent road collapse, by utilizing fiber optic modulation and demodulation equipment from operators and optimizing algorithms to monitor road voids. In recent years, urban road collapse disasters have occurred in some areas. Starting from 2022, the center has been developing distributed fiber optic sensing technology, which connects existing fiber optic modulation and demodulation equipment to receive and analyze vibration signals restored by fiber optic scattering light, thereby providing technical support for eliminating road hazards. "We hope that through a single fiber optic cable, underground cavities on roads can be monitored 24/7." Assistant Director Wu Dong said in an interview with Science and Technology Daily. "The fiber optic sensing overlay algorithm can detect different" symptoms "of roads. Currently, ground penetrating radar is the main means of detecting underground cavities. This method can accurately detect the position, depth, and size of underground cavities. However, in practical use, this method has high procurement and service costs, slow detection speed, and is difficult to promote on a large scale due to the influence of groundwater." Wu Dong introduced that there is an urgent need to find a monitoring and perception network that can expand the monitoring range of sensors, reduce the cost of sensor deployment, and dynamically perceive the vibration, temperature, and strain of urban infrastructure in real time, accurately, and dynamically. And optical fiber is the preferred choice for center researchers. When light propagates in optical fibers, it undergoes changes due to the state of the fiber and the surrounding environment of the fiber. By utilizing principles such as Rayleigh scattering and Brillouin scattering, vibration, strain, and temperature around optical fibers can be measured, analyzed, and located. Specifically, it means connecting optical fibers to fiber optic modulation and demodulation equipment. Due to the microscopic non-uniformity of optical properties such as density and refractive index of optical fiber materials, light will scatter when propagating in optical fibers Wu Dong introduced that when the environmental temperature changes or the optical fiber undergoes deformation, the phase of scattered light in the optical fiber will change accordingly. After the light signal is scattered back to the fiber optic modulation and demodulation equipment, it will be restored to a vibration signal. These vibration signals each have their own characteristics. We have built a "voiceprint library" for these specific scenarios, where the vibration, strain, and temperature signals of the optical fibers will vary when roads are hollowed out, subjected to destructive construction, or when rain and sewage pipelines are structurally damaged. The computer algorithm in the background can determine what problem is on the road by comparing it with the 'voiceprint library' Wu Dong gave an example that when a vehicle passes through a road, each sampling point on the fiber optic can collect vibration signals of the vehicle passing through. However, the signals at underground cavity points have greater attenuation compared to normal point signals. Therefore, analyzing fiber optic vibration signals can effectively monitor road cavities. At present, the fiber optic vibration analysis algorithm developed by the team can identify job behaviors such as excavator smashing, air cannon operation, cutting machine operation, compactor operation, and pipe jacking machine operation. Road experiments were conducted to verify the feasibility and scientificity of the monitoring plan, with a monitoring accuracy of 85%. In 2023, the research team selected a 2-kilometer road on Yudao Street in Nanjing for experimental testing and verification. The results showed that there were densely distributed small-scale voids on the outer lanes of both sides of the road, with a minimum void diameter of about 1 meter, and an estimated total of more than 30 voids. After comparing the test results with the municipal road maintenance data, we found that the overall monitoring accuracy reached 85%. Of course, the voids are mainly municipal manholes. Nanjing is not a city with a high incidence of road hollows Wu Dong said. The verification results of construction monitoring have also greatly inspired the research team. On November 2, 2022, distributed fiber optic sensing detected that from 13:36 to 13:42, there was a construction machinery crushing operation near the entrance of the parking lot on the east side of Nanjing Deji Plaza Phase II. The project team members immediately rushed to the entrance of the parking lot to confirm and found that there was indeed excavator operation inside the fence. If you want to sense road hazards, the denser the selected monitoring points, the better. Wu Dong introduced that distributed fiber optic sensing technology can currently achieve continuous perception information acquisition for 40 kilometers, with risk positioning errors within the meter level. Nowadays, the team sets the interval distance between conventional fiber optic monitoring points at 10 meters. Once signal abnormalities are detected, the monitoring points can also be encrypted to more accurately identify road hazards. More importantly, by utilizing distributed fiber optic sensing technology, there is no need for road breaking construction, and monitoring can be carried out by simply connecting the fiber optic cables in the existing communication cables of the operator to the sensing equipment. Generally, upgrading the hardware equipment deployed in the data center as needed within 3 to 5 years is sufficient, which is very easy to maintain Wu Dong said that currently, the team has deployed distributed fiber optic sensing equipment and debugged algorithms in Wuxi, Kunshan in Suzhou, Suqian and other places. Next, they will conduct regional demonstrations to continuously improve the accuracy of identifying adverse events such as road voids, while reducing usage costs. In addition to urban roads, distributed fiber optic acoustic sensing systems can also be used in other scenarios. Wu Dong introduced that urban rainwater and sewage pipes often suffer from structural damage. Existing technological methods generally require emptying and flushing the pipes, allowing pipeline robots to enter and take photos, and then manually judging whether there are "diseases", which is inefficient and costly. Relying on a distributed fiber optic acoustic sensing system, the vibration signal propagated along the pipeline is received through pipeline optical fibers. When encountering structural defects such as disconnection, virtual connection, and misalignment, the vibration characteristics will change. Therefore, by extracting vibration features, the location of "diseases" can be identified. "In the future, the distributed fiber optic acoustic sensing system can be equipped with multiple algorithms such as construction operation and equipment type identification, road cavity monitoring, and rain and sewage pipeline detection, meeting the multi scenario business needs of urban infrastructure safety operation monitoring, and safeguarding urban safety." Wu Dong said. (Lai Xin She)