Massive Data Capture Approach for Modeling Existing Building Stocks

This research paper aims to develop an approach for the digitalization of non-heritage building stock. Existing building stocks in need of rehabilitation are still not subject to optimized, massive digital surveying processes. Thus, it is difficult to assess the performance of the stock in its current state and after potential retrofitting. While massive data capture is being used to model heritage cases with high precision for preservation and documentation projects, this research paper aims to develop an approach for the digitalization of non-heritage building stock that allows for broader implementation, quicker results, and higher scalability, reducing the time required for data capture but still being precise enough for rehabilitation processes. The novel approach combines a laser scanner, thermal infrared sensing, high-quality pictures (HQPs), and automatic frame extraction (AFE) from video. Data preparation for three-dimensional reconstruction is the main novelty of this approach, which has been validated to obtain the surroundings and building information model (BIM) of the reference building for Barcelona schools. The results coincide with previous projects regarding the high precision of a laser scanner and the coverage of photogrammetry. New findings indicate that HQPs are a highly efficient method. Its combination with AFE provides higher levels of coverage. The proposed approach moves forward from the manually modeled BIM misalignments and enables modeling entire clusters to obtain digital twin building stock to ease future management of existing buildings.