Digital twins to save our steel bridges

The growing number of aging and vulnerable bridges has highlighted the need for effective maintenance methods to increase safety and decrease maintenance costs. To address this growing need for efficient bridge maintenance, this thesis explores the integration of digital twin technologies into existing infrastructure systems. By analyzing current challenges in bridge maintenance, this study aims to demonstrate the potential of digital twin models in enhancing maintenance efficiency and accuracy. Utilizing a Design Science Research Methodology, this work constructs baseline and target architectures, providing a comprehensive view of potential improvements. This offers a structured approach towards adopting digital twin technologies in infrastructure management, alongside a detailed migration strategy and implementation roadmap. To validate the proposed architecture, a prototype digital twin is developed for a steel bridge, with its feasibility demonstrated through an experimental implementation on a real-world steel bridge. In addition, a second case study is conducted to illustrate an alternative implementation approach of the proposed architecture. Together, these two variations demonstrate the flexibility and applicability of the design while also identifying future research directions for its practical implementation and comprehensive evaluation. To evaluate the system, simulations are conducted in Simulink, Uppaal, and GeNIe for time-based analysis, state machine analysis, and value of information analysis using Bayesian decision modeling, respectively. Additionally, a virtual reality application developed in Unity is used to explore remote bridge inspection and to assess the impact of augmented reality and artificial intelligence in enhancing inspection through a controlled experiment. This study not only contributes to the field of infrastructure management but also serves as a guide for future digital transformation initiatives in similar sectors. Digital twin technologies can be employed to revolutionize bridge maintenance and integrate other new technologies, such as artificial intelligence and extended reality. Moreover, the digital twin can standardize different assessment systems and models, ensuring interoperability. This study takes the first step in realizing this digital transformation by providing digital twin-enabled reference architectures for bridge assessment and maintenance.