Autonomous Robotic System for Pipeline Integrity Inspection.

In this paper, we present an external pipeline inspection robotic system capable of detecting the surface defects on above-ground pipelines and modeling their degradation. This system consists of two subsystems, the main base and the sensing system. The main base is a self-driving autonomous ground vehicle (AGV) equipped with Lidar, acoustic sensors and motor encoders, which can track the pipeline on uneven terrains. The sensing system includes two cameras attached to a C-arm, which rotates around the pipe. The two cameras are placed 180° from each other and can take pictures at 90° intervals, allowing the system to analyze the full 360° of the pipe with only half of the rotation as compared to a single camera system. When the C-arm encounters a flange or support, it retracts off the pipe, moves past the obstacle, and extends back to continue taking images. These images are used for defect detection. The detected defect data is then used for modeling the defect degradation and predicting when the defects become critical and require maintenance.