Abstract
Condition monitoring and defect inspection in the buried oil and gas pipelines, made of ferrous material, has always been a challenge for all organizations operating in the Oil & Gas sector. Pipelines can be inspected in two ways, internally and externally. Internal inspection by ILI tools require special infrastructures like pig launchers and receivers along with pre-preparation before inspection like internal cleaning there is the data collecting and analysis process which is time-consuming. Whereas in communally used external inspection a group of workers drive a vehicle along the pipelines to perform visual inspection of the pipelines for detection of leakage or any other kind of visible damages. Such manual external inspection is highly inefficient, expensive and hazardous. In such a way it is difficult to obtain any important information for the anomalies brewing in the buried pipes or cathodic protection layer.
A lot of work has been done towards developing NDT technologies to inspect pipelines. However, most of the NDT sensors work only in close vicinity of the pipeline surface which requires an excavation of the pipelines and exposing the structure. This shortcoming of NDT techniques has attracted researchers towards other NDT techniques such as non-invasive magnetomatric diagnosis (NIMD) which allows non-contact detection of anomalies from distance in the core metal of the pipelines deeply buried underground. NIMD sensors work on principle of measuring distortions of residual magnetic fields conditions by the variation of the pipeline’s metal magnetic permeability in stress concentration zone due to combined influence of various factors such as residual stress, vibration, bending and loading of pipelines, installation stress and temperature fluctuations etc. These handheld magnetic sensors are used manually by field operators therefore inspection of long pipelines in extreme environmental conditions is not feasible. A non-contact external robotic inspection system, Autonomous Ground Vehicle (AGV), carrying such non-contact magnetic and visual sensors is designed and tested in this work.
AGV is equipped with two kinds of sensors, the first navigation sensors and the second inspection sensors. Accurate autonomous tracking of the pipelines by the AGV is achieved by fusion of three navigation mechanisms based on visual data, GPS and pipe locator. The pipe locator in combination of CP post is one of the most extensively used sensors in the oil and gas industry for tracking the buried pipelines. In this work manually used pipe locator is now fully automatized for autonomous tracking of the buried pipelines by the AGV. For this purpose, a hybrid automata trajectory controller is developed for a non-holonomic AGV where a PID controller is combined with non-linear forward velocity of the AGV depending on the lateral distance error and angular alignment error. Field experiments are conducted successfully to demonstrate accuracy of the newly designed controller. Successful development of such complex mechanism requires solution of many critical challenges like teleoperation, system and supervisory controls, trajectory tracking, image processing and sensor data fusion.