Automated 3D BIM modeling for Digital Twin development

This project focuses on fully automatically 3D BIM modeling using 3D point cloud data. In this project, we firstly collected 3D point cloud data of small room by using iPhone 12 Pro (integrated LiDAR sensor). The collected data then will be treated as input data of 3D point cloud recognition model. This deep learning model was trained on both public dataset (ScanNet V2) and our private dataset.
For constructing 3D BIM model from predicted results, we developed an application based on REVIT APIs (C#) that can automatically construct 3D BIM model of room inside objects such as wall, floor, ceiling, tables, chairs, cabinets, etc. These 3D BIM models of the room can be integrated with our Digital Twin System or with IoT sensor database for simulation.