Autonomous operation of a robot dog for point cloud data acquisition of scaffolds

Scaffolds are essential temporary structures on construction sites. Since scaffolds are frequently installed and dismantled, the inspection needs to be performed in real-time. This paper proposes a framework to automate the acquisition process of scaffold point cloud data using a robot dog. First, a Simultaneous Localization and Mapping (SLAM) algorithm (LIO-SAM) is deployed for real-time map creation based on laser-based 3D data. Scaffolds are automatically detected using the bird's eye view (BEV) projection images of the registered 3D point clouds. A scanning distance is also determined for each detected scaffold to move the robot dog to an optimal location. The robot dog can successfully scan the scaffolds on construction sites by using the proposed framework.