Integrating Building Information Modeling, Deep Learning, and Web Map Service for Sustainable Post-Disaster Building Recovery Planning

Building recovery is always a major issue after disasters due to the urgent need for efficient assessment of damaged buildings and effective management of required resources. At the same time, it is necessary that building recovery planning considers the impact of used materials on the environment as well as the proper allocation of budgets. In this paper, we propose a novel integrated post-disaster building recovery planning system that utilizes Building Information Modeling (BIM), Deep Learning (DL), and Web Map Service (WMS). BIM provides 3D models and detailed information for buildings; DL enhances the accuracy and efficiency of building assessment; WMS generates reliable data on locations and for route planning. The BIM-DL-WMS integration would contribute significantly to building recovery planning by automating building element assessment, cost estimation, and carbon emission analysis for repair/retrofit materials and material delivery. This developed system would make post-disaster building recovery planning more efficient, effective, and sustainable. The framework of the BIM-DL-WMS system is elaborated on in detail in this paper. The implementation of the system could be applied to not only single-building recovery planning but also the recovery planning of a district or even a town or city.