SCAAI tool
With the integration of Artificial Intelligence in the construction sector, digitalization has steadily grown towards the adaptation of innovative technologies in all buildings’ life cycle. In Inperso the application of AI to the scanning activities has brought to the development of the SCAAI tool, by R2M.
Scan-to-BIM technologies generate 3D models to interpret, manage and support the overall process of both building renovation and operational phase. Using AI to capture and recognize assets and enable the enrichment of models with information stored in digital BIM libraries, foster the process of scanning and then modeling buildings and construction sites. The overall process can be summarized in 5 steps: first scanning the environment through a LIDAR technology - like the 360° Matterport camera -,
second the postprocessing of the point cloud obtained in the first phase takes place, third the point cloud obtained is processed through a web platform AI based in which the BIM model is created, fourth objects can be detected from the point cloud and suddenly added to the BIM models, fifth the BIM model is finally uploaded to the RE-Suite platform. Therefore, in combining different technologies, INPERSO is accelerating data collection and modeling processes in the AEC sector.
Objectives
SCAAI aims to develop a fast automatic detection of building objects from the point cloud generated using scanning technologies, thus reducing the scanning onsite timing and creating a BIM model that could be further used for design purposes once uploaded to the RE-Suite platform.
Work Carried Out So Far & Milestones Achieved
During the Inperso project development, the SCAAI tool application path has been divided into two sections, firstly the scanning activities at pilot sites using the Matterport Camera °360 which results in 3D Virtual Models and a point cloud that were used in the second step of the tool’s development. Therefore, in the second section, the working group researched and evaluated different approaches and methods to postprocess the point-could using AI.
The results lead to the use of point cloud segmentation to detect objects. Lastly this automatic scan2BIM process was integrated into a web service using APIs to upload the output of the process to the RE-Suite platform. The work was performed according to project outcomes achieving the expected results.
Remaining Work
The technology has been developed by R2M and finalized in December 2023. No further activities in technology implementation are planned until the project completion.
Expected Impact
The expected impact of SCAAI in the construction sector relies on the reduced percentage of timing in the scan2BIM process. A concise support from AI can support simple BIM model creation through 3D point cloud from laser scans more quickly, reducing scanning time by over 80%.