Bus network based fleet monitoring towards sustainable transport infrastructure

Since its first invention in 2004, drive-by structural health monitoring has been widely adopted to assess ageing infrastructures worldwide. In 2019, the European Union Joint Research Centre highlighted it as one of the most promising techniques for bridge monitoring. Recently, a fleet composed of different vehicles has been regarded to be more efficient in obtaining bridge information in contrast with a sole vehicle, as fleet monitoring can mitigate the annoying impact caused by undesired road components, and any uncertainty from a single passing vehicle. Conventional drive-by fleet monitoring builds on instrumented heavy trucks, whereas various mechanical properties of the tested trucks usually make it challenging to derive the bridge information for some scenarios, i.e., both the dominant frequency of the vehicle and the bridge located within the same range. Therefore, this paper proposes a novel drive-by fleet monitoring framework, where thousands of buses are generated by utilizing Monte Carlo Simulations. Latterly, bridge information, i.e., operating deflection shapes, is extracted by implementing the proposed algorithm from the measured bus accelerations for bridge damage detection and localization. Unlike the conventional drive-by monitoring, bus network-based fleet monitoring has proved to be more efficient because most of the mechanical properties of buses are identical, which further contributes more benefits in bridge information extraction. The proposed bus network-based drive-by fleet monitoring is validated with numerical experiments of a short-span concrete bridge.