Scour Monitoring Method Using Ultrasonic Based Flow Meter and Advanced Machine Learning
Bridges play a critical role in emergency responses when natural disasters strike. However, scouring is the primary cause of bridge failures and as a result is of major concern during natural hazards such as floods. As such, the use of more advanced technologies for Bridge Scour Monitoring (BSM) would lead to major improvements in many aspects of emergency management. Such technologies would also help improve maintenance scheduling, simultaneously making the infrastructure system more economically efficient and technically reliable. In this paper, we propose a quasi-real-time scour monitoring method which is a result of the development of a device that consists of ultrasonic flow meters and using One-Class Support Vector Machines to detect scour and classify the levels of both scour and scour refill.