Reality Capture of Hydraulics Infrastructure

Abstract
Topographic surveys of large culvert infrastructure have typically required significant financial investments, complicated logistics, and extensive data processing, all carried out by highly trained staff. More recently, technologies such as Light Detection and Ranging (LiDAR) – which measures the distance from a sensor using the Time-of-Flight of a light pulse - and multi-camera smartphones have reduced the costs and complexity needed to obtain high-resolution, 3D models of obscure built infrastructure.

In the summer of 2022, the New Brunswick Department of Transportation and Infrastructure (NBDTI), Design Branch partnered with Modelar Technologies to evaluate the potential use of LiDAR-enabled iPhones as a tool for capturing 3D models of the interior of large culverts. The scanning software fuses multiple HD images, spatial orientation data and distance measurements from the iPhone’s LiDAR sensor to construct both a 3D point cloud and a textured, triangulated 3D surface model. The quality of the resulting 3D models depends on a number of factors, including the speed at which the scanning device is moving, the sensor resolution and the lighting conditions inside the culvert.

This paper presents various aspects of this new reality capture collection system and compares the speed and cost to traditional surveying and data collection methods. The advantages (and possible drawbacks) of data collection by a wider range of staff at relatively lower cost than traditionally possible are explored. The accuracy and precision of these models are explored. Merging high-precision survey data with the mobile reality capture data and UAV 3D model to create a digital twin rendering of the asset will be shown. Additionally, the long-term goals for end-to-end processing workflows and data management are presented.

Keywords: digital twins, reality capture, scanning, SLAM, 3D modelling, structures, culverts