On August 9, during the inaugural visit to Mainz University of Applied Sciences by Clemens Hoch—who has served as Minister of Science and Health for Rhineland-Palatinate since May 2021—the Department of Engineering presented a “PhD Student Standup” on the forward-looking topic of Building Information Modeling (BIM).
BIM is a digital technology used to describe and represent information related to the planning, design, construction, and operation of buildings. BIM increasingly describes all aspects of the built environment, including civil infrastructure, utilities, and public spaces. This allows various datasets to be displayed within the same information environment and analyzed efficiently.
All three disciplines within the Department of Engineering—Architecture, Civil Engineering, and Geoinformatics and Surveying—have made this forward-looking topic a priority in both teaching and research. In addition to ongoing doctoral dissertations on the topic, interdisciplinary projects are being conducted by the departments of Architecture and Civil Engineering on the integration of planning processes through BIM.
With the help of augmented reality and machine learning, we can classify 3D point clouds directly during data capture using a smartphone. Graphic: Bastian Plaß, CC BY-SA 4.0
Bastian Plaß from i3mainz is focusing his dissertation on process automation in the digitization of existing buildings using self-learning software solutions from the field of machine learning (Scan-to-BIM). His focus is on structuring 3D point clouds resulting from building surveys. In addition to established methods from the fields of semantics and deep learning, Plaß is also testing new approaches, such as real-time segmentation using a smartphone app. Using augmented reality and machine learning, the 3D data can be structured in real time, which significantly simplifies the established Scan-to-BIM process and thus opens up new areas of application. The iOS prototype developed is versatile and is currently being tested and further developed as part of the BIM4cAIre project.
