Long-term monitoring of bridge displacement using LiDAR, computer vision, and strain-based indirect estimation

Abstract

Long-term bridge deflection, which is generally caused by concrete creep, shrinkage, structural damage, and loss of tendon forces, can be used as an indicator of structural integrity and safety. Nonetheless, the long-term deflection has been rarely used for bridge maintenance purposes in practice, because measuring bridge deflection for a long time is quite challenging particularly in full-scale bridges. Conventional deflection measurement approaches have limitations for long-term measurement. A typical way is to use Linear variable differential transformer (LVDT) with a temporary supporting structure such as scaffold, which can be a source of substantial error. Global Positioning System (GPS) and computer vision-based approaches, which have been recently studied, have critical issues related to measurement accuracy and equipment movement in the long-term measurement. Indeed, a feasible means for long-term displacement measurement is desired. This study introduces three different types of long-term displacement measurement approaches: (1) Light Detection and Ranging (LiDAR), (2) strain-based indirect estimation, and (3) computer vision method using a double camera system. These methods are carefully developed to handle possible long-term errors such as sensor movement. The three displacement measurement methods are applied to a railway bridge in Korea, acquiring displacement for about six months in the construction stage. The performance, advantages, and disadvantages of each method learned from the experiment are discussed.