Algorithms for autonomous subsea pipeline tracking
MetadataShow full item record
One of the major benefactors of unmanned submersibles is Oil & Gas industry for applications in inspection of subsea pipelines, where an Autonomous Underwater Vehicle (AUV) is required to follow the target pipe over long ranges. This is an expensive and time consuming process when done using combination of tethered Remotely Operated Vehicles (ROV), divers and surface vessels with large crew. Even with some AUVs this can take considerable time. So current research aims at lowering down the inspection costs by using new sensors and algorithms onboard the AUV that can conduct high fidelity survey faster than current methods. The aim of this project is to develop detection, tracking and sensor fusion algorithms for subsea pipeline tracking. The individual sensors proposed for tracking complements each other well to solve both visual and buried pipeline tracking problem. For exposed pipeline tracking a combination of multibeam echosounder and camera is proposed and for buried detection, a magnetometer is suggested. This work puts focus on developing a sensor fusion methodology for tracking subsea pipelines using 3-axis magnetometer, a multibeam echosounder and an optical camera. The proposed method has potential to track buried pipelines more efficiently without losing track of target pipeline, thus it brings the overall cost down from reduction in re-surfacing requirements for a GPS fix in case of loss of pipeline track. When the pipeline is partially buried, a combination of multibeam echosounder and camera-based segmentation works well as camera detections are better in such a scenario. Tracking solely using a magnetometer is not feasible due to nature of pipeline’s magnetism. Thus, it is used in conjunction with other two sensors to track pipeline when it is buried for short distances to avoid losing track. Performance of detection algorithm is tested experimentally for individual sensors and then fused into a single Bayesian framework to produce a robust subsea pipeline tracker that can be used by the vehicle controller for guidance along the pipeline.