Deriving Digital Twin Models of Existing Bridges from Point Cloud Data Using Parametric Models and Metaheuristic Algorithms

In building information modeling (BIM), a digital twin (DT) is a model that represents the current status of an existing structure; thus, facilitating the operation and management process. Due to higher measurement speed and accuracy, laser scanning and photogrammetry are generally employed, resulting in point cloud data (PCD). Today, the required volumetric models are created in a laborious and costly manual process from PCD. This paper aims to automate this process by applying metaheuristic optimization algorithms to fit highly parametric BIM models of bridges into given point clouds. For this purpose, parametric base models of elements are created and instantiated by adjusting their parameters' value using metaheuristic algorithms. This optimization process leads to extracting the parameters for a model from PCD and creating 3-D volumetric shapes. The paper's results show that metaheuristic algorithms can be successfully used for parametric modeling even in point clouds with occlusion and clutter.