Planning for Scanning Using Building Information Models: A Novel Approach with Occlusion Handling

Terrestrial laser scanning is finding an increasing range of applications in the Architectural Engineering Construction and Facilities Management (AEC/FM) industry. While significant progress has been made in the performance of laser scanners and multi-scan registration, planning for scanning -- i.e. the selection of locations for the scanner and registration targets -- is still done quite subjectively by surveyors and is underpinned by little and basic scientific reasoning. This may lead to 3D point cloud data being incomplete or insufficiently accurate to deliver the completeness and accuracy expected from the subsequent measurement or modelling tasks.

In this paper, preliminary results are presented for a novel scientific approach for planning for scanning in the construction sector. The approach is designed to generate automatic laser scanning plans using as input: (1) the facility's 3D BIM model; (2) the scanner's characteristics; and (3) the scanning specifications in terms of individual point precision and surface area covered by the scanned data for each 3D BIM model object. The output is the smallest set of scanner locations required to achieve those requirements. The particular value of the proposed approach is its capacity to take model self-occlusions into account. The performance of this approach is assessed with a simple experiment simulating the scanning of a concrete structure.