A conical laser light-sectioning method for navigation of autonomous underwater vehicles for internal inspection of pipelines

This paper presents a novel sensing method for navigation of Autonomous Underwater Vehicles, AUVs, through pipelines to conduct autonomous internal inspection. Unlike remotely operated pipe inspection robots, AUVs do not have an umbilical cable and so they can easily maneuver through bent sections of pipelines and the distance they can cover is not restricted by the length of the cable. Presently pipe inspection robots come in contact with the walls of the pipe during their operation. However old and aged pipelines may have loose corroded materials or biological growth which may get detached and pollute the fluid or further damage pipe interior when the pipe inspections robots touch the walls. AUVs can operate without coming in contact with the pipe wall and so this technique is a non-contact measurement and inspection technique. The proposed navigation sensor makes use of computer vision techniques to estimate the relative position and orientation of the vehicle inside the pipe with 4 degrees of freedom, which will enable the AUV to swim through the center of the pipe. A conical laser is projected on the pipe wall and the image of the laser is acquired by a camera. The features of the image are extracted and matched with a feature database prepared apriori. The position and orientation of the matching feature record in the database gives the estimated position and orientation of the vehicle. Experiments were conducted in dry experimental pipelines to verify the performance of the proposed sensor and results are presented. Control Simulations were performed to verify the ability of the sensor to navigate an AUV. The results of these controls simulated are also presented in this paper.