

A Digital Twin of Bridges for Structural Health Monitoring
Abstract
Bridges are critical infrastructure systems connecting different regions and providing widespread social and economic benefits. It is therefore essential that they are designed, constructed and maintained properly to adapt to changing conditions of use and climate-driven events. With the rapid development in capability of collecting bridge monitoring data, a data challenge emerges due to insufficient capability in managing, processing and interpreting large monitoring datasets to extract useful information which is of practical value to the industry. One emerging area of research which focuses on addressing this challenge is the creation of ‘digital twins’ for bridges. A digital twin serves as a virtual representation of the physical infrastructure (i.e. the physical twin), which can be updated in near real time as new data is collected, provide feedback into the physical twin and perform ‘what-if’ scenarios for assessing asset risks and predicting asset performance. This paper presents and broadly discusses two years of exploratory study towards creating a digital twin of bridges for structural health monitoring purposes. In particular, it has involved an interdisciplinary collaboration between civil engineers at the Cambridge Centre for Smart Infrastructure and Construction (CSIC) and statisticians at the Alan Turing Institute (ATI), using two monitored railway bridges in Staffordshire, UK as a case study. Four areas of research were investigated: (i) real-time data management using BIM, (ii) physics-based approaches, (iii) data-driven approaches, and (iv) data-centric engineering approaches (i.e. synthesis of physics-based and datadriven approaches). A framework for creating a digital twin of bridges, particularly for structural health monitoring purposes, is proposed and briefly discussed.
DOI
10.12783/shm2019/32287
10.12783/shm2019/32287