Development Of Technology For Automatic Extraction Of Architectural Plan Wall Lines For Concrete Waste Prediction Using Point Cloud

Taehoon Kim Kyungpook National University
Soonmin Hong Kyungpook National University
David Stephen Panya Kyungpook National University
Hyeongmo Gu Kyungpook National University
Hyejin Park Kyungpook National University
Junghye Won Kyungpook National University
Seungyeon Choo Kyungpook National University

Recently, as more and more projects on residential environment improvement in cities are actively carried out, the cases of demolishing or remodelling buildings has been increasing. Most of the target buildings for such projects are made of concrete. In order to reduce energy use as well as carbon emissions, the amount of concrete used as a building material should be reduced. This is because the concrete is the largest amount of construction waste, which the exact amount of concrete needs to be predicted. The architectural drawings are essential for the estimation and demolition of building waste, but the problem is that most of the old buildings’ drawings do not exist. The 3D scanning process was performed to create the plans for such old buildings instead of the conventional method that is long time-consuming and labour-intensive actual measurement. In this study, we scanned 40 old houses that were scheduled to be demolished. The result showed that the 3D scanned drawings’ accuracy – 99.2% – was higher than the ones measured by the conventional way. Through the algorithm developed in this study, the various processes of demolition, drawing measurement, and discarding quantity prediction can be solved in one process, thereby reducing work efficiently. And, considering the reliability of the research results, it is possible to reduce the economic loss by predicting the exact amount of waste in advance. After that, if the algorithm, developed in this study, can be further subdivided and supplemented to identify the materials for each part of the old buildings, it will be able to propose an efficient series of processes that distinguish between recyclable materials and wastes and thereby efficiently dispose of them.

Keywords: Point Cloud, Construction Waste, Parametric Design, Algorithm, Automatic Extraction, Un Sdgs 8(Decent Work And Economic Growth)

View Full Paper