To address the need for a shift from a linear to a circular economy in the built environment, this paper develops a semi-automated assistive process for planning building material deconstruction for reuse using sensing and scanning, Scan-to-BIM, and computer vision techniques. These methods are applied and tested in a real-world case study in Geneva, Switzerland, with a focus on reconstruction and recovery analysis for floor beam systems. First, accessible sensing and scanning tools, such as mobile photography and smartphone-based consumer-grade Lidar devices, are used to capture imagery and other data from an active demolition site. Then, photogrammetry and point cloud data analysis are performed to construct a 3D BIM model of relevant areas. The structural relationships between reconstructed BIM elements are evaluated to score the feasibility for recovery of each element. This study illustrates what is feasible and where further development is necessary for automating building material reuse planning at scale to increase the uptake of circular economy practices in the construction sector.@en

The integration of repurposed material in new construction products generates resiliency strategies that diminish the dependency on raw resources and reduce the CO2 emissions produced by their extraction, transportation, and manufacturing. This research emphasizes the need to expand preliminary data collation from pre-demolition sites to inform early design decisions and proposes a support system for its implementation. Material (data) Intelligence investigates how the merging of artificial intelligence and data analysis could have a crucial impact on achieving widespread material reuse. The first part consists of automating the process of detecting materials and construction elements from pre-demolition sites through drone photography and computer vision. The second part links the resulting database with a computational design tool. This paper strengthens the potential of circular material flows in a digital paradigm and exposes the capability for constructing big data sets of reusable materials, digitally available, for sharing and organizing material harvesting.@en