Effectiveness of Drone-Based Photogrammetry for On-Site Quantity Assessment

Abstract

Drone-based surveying using laser scanning with LIDAR has been routinely used in heavy construction to produce high-quality topographical data with increased grid/mesh density when compared to traditional surveying. Differences between pre-construction conditions and post-construction topographical scans thus provides a basis for identifying the work performed, measuring progress and managing progress payments. The technique has demonstrated cost-effectiveness with less physical effort and less time becoming the go-to technology on large unit-price heavy construction projects. The approach is not routinely used on smaller projects partially due to the cost of acquiring the technology or outsourced services. Alternative technologies are desired which could benefit smaller and potentially more diverse construction activities. One promising approach involves using reality capture based on high-resolution imagery. This approach processes a cloud of geolocated imagery to develop virtual models which can then be integrated and analyzed through a BIM platform. Drones-based imagery through off-the shelf unmanned autonomous vehicles (UAV’s) provide the ideal input for this reality capture techniques. The reliability of the resulting models is unknown, however, and research was performed to identify the reliability of the technique to determine on-site volumetric and area quantities. Various off-the-shelf UAV’s (drones) were evaluated and statistical techniques were employed to quantify the reliability of the resulting models. Regression was used to extract data obtained through systematic trials to site-level applications. Results are presented and future directions are outlined and discussed. Diamond Congress Ltd. Peer-review under responsibility of the scientific committee of the Creative Construction Conference 2019.