A Digital Twin Framework for Bridges

Current inspection and maintenance practices for bridges involve periodic manual visual surveys which are time consuming and cumbersome due to number of bridges in need of assessment. Unprecedented capabilities have become possible in recent years thanks to strides made in sensing technologies, computer vision, and machine learning. These technologies can be integrated into a digital twin framework for better insight into a bridge’s structural integrity, maintenance needs, and potential risks. A digital twin framework involves creating a virtual model that mimics the behavior and performance of a physical bridge. Data collection about a bridge’s physical condition is the first step in developing a framework. Physics simulation and graphical environment are next needed to host relevant models of the bridge. An illustrative example involving these steps and digital twin use for scenario simulation for different mechanisms is provided. The example focuses on modeling of a railway bridge in Lyndhurst, New Jersey.