Home | Contact ST  
Follow ST

Feature Articles

Pipeline Inspection With Interferometric SAS
New Sensors and Techniques Converge to Allow More Cost-Effective Inspection Using AUVs

By Per Espen Hagen
Senior Principal Engineer
Dr. Even Børhaug
Research and Development Engineer
Kongsberg Maritime AS
Horten, Norway

Øivind Midtgaard
Principal Scientist Norwegian Defence Research Establishment
Kjeller, Norway

External inspection of underwater pipelines is an important and time-consuming task. The objective of these inspections is to detect burial, exposure, free spans and buckling of the pipeline, as well as possible damage due to trawling, anchoring and debris near the pipeline. Today, inspection is generally performed with towed or remotely operated vehicles (ROVs) and specially equipped vehicles rolling on top of the pipeline.

Technology is now available that can potentially make pipeline inspection with autonomous underwater vehicles (AUVs) a very cost-effective alternative to much of the ROV-based inspection. AUVs are more stable platforms that can travel faster (typically three to five knots, compared to one to two knots for most ROVs), and they can perform their task without being followed closely by a large surface vessel.

Traditional sensors for pipeline inspection are side scan sonar, multibeam echosounders and video cameras. These sensors require different geometries, meaning that several passes must be made in order to get a complete data set. Also, with each of these sensors, the pipeline must be detected and tracked in real time on board the AUV, using data from the inspection sensor itself. This is because these sensors typically have a swath (or at least a 'sweet spot') that is narrower than the sum of the navigational uncertainty and the uncertainty of the pipe-lay position. The tracking software must be able to cope with pipe burial scenarios where multiple pipelines are visible on a wide variety of seafloor types—from flat sandy bottom to cluttered, rocky seabed. For inspection with a video camera, the AUV must be positioned directly over the pipeline at very low altitude.

By equipping an AUV with a high-resolution interferometric synthetic aperture sonar (SAS) and a camera, data quality and area coverage can be improved significantly. In a joint research and development effort, Kongsberg Maritime AS and the Norwegian Defence Research Establishment (FFI) are developing a new concept for AUV-based pipeline inspection, with partial funding from the Research Council of Norway. The system will be demonstrated on a HUGIN 1000 AUV in 2010 and 2011.

HISAS single-ping sector-scan imagery with two different pipelines near its maximum range of 185 meters. A standard side scan sonar would give one beam along the center of the sector (illustrated to the right of each sector scan).
The basic concept is as follows: An AUV with a high-resolution interferometric SAS (in this case, the Kongsberg HISAS 1030) and an optical sensor is launched from a support vessel. The AUV is programmed to follow a route 50 to 100 meters to the side of the as-laid position of the pipeline. HISAS provides a constant-resolution swath of 150 to 200 meters on each side of the pipeline. Because of the wide swath and the high navigation accuracy of HUGIN, prior knowledge about the pipeline position does not have to be very precise. The pipeline is detected and tracked in real time using side scan or sector scan data from HISAS together with prior position data. These data will be used to generate updated pipeline positions.

The AUV can survey 20 to 30 nautical miles of the pipeline in this way while the support ship performs other tasks (e.g., ROV-based intervention or maintenance on previously surveyed sections). HUGIN then turns 180 degrees and returns to the launch position.

On the way back, it may travel at the same distance to the other side of the pipeline in order to cover an even wider corridor. Alternatively, the AUV can travel back directly above the pipeline, using the pipeline position data generated during the first half of the mission. In that case, HUGIN changes altitude to five to 10 meters above the pipe and turns on its optical still image camera and light-emitting diode (LED) strobe to record high-resolution optical imagery of the pipeline. In addition, a multibeam echosounder is used to generate bathymetry of the pipeline and its immediate surroundings. The data are used to track the pipeline in real time and adjust the AUV path as needed to ensure optimal imaging geometry.

Pipeline burial is detected and mapped on the first half of the dive, so the AUV does not have to perform time-consuming searches when passing over buried pipeline sections during the optical mapping phase.

Detection and Tracking
HISAS 1030 functions as a dynamically focused side scan sonar in real time, providing imagery suitable for pipeline detection and tracking. A detection and tracking module is under development at FFI and will be used during sea trials with the system later this year. The detection module applies a nonlinear matched filter to identify pipe-like responses in each sonar ping. A Kalman-filter-based tracker compares new detections to the predicted positions of existing tracks. Tracks with an associated detection are updated while new tracks are created for leftover detections. Persistent tracks are reported to the inspection-control module as candidate pipeline paths.

The reported tracks are compared with a priori information, e.g., pipeline as-laid data, to ensure that the correct pipeline is selected even if multiple pipelines are visible within the sonar swath. Also, for buried pipe sections, a priori data will be used to extrapolate the pipeline position estimate from the last detection.

Data from HUGIN 1000/HISAS 1030 pipeline survey experiments in 2009 show that pipelines can give very strong returns (due to specular reflection) when the AUV travels parallel to the pipeline, but the echo may be barely discernible at other aspect angles. The shadow will still give away the location of the pipeline, but it must then typically be followed for some time to provide a high-confidence track.

HISAS imagery from a pipeline survey showing
a 100-by-200-meter section of the starboard
swath (range 45 to 145 meters). The inset
shows a 10-by-10-meter cutout around an
anchor found close to the pipe.
One of the advantages of HISAS is that sector-scan imagery can be generated from each ping: HISAS insonifies over a 20-to-40-degree range of azimuth angles, and beams can be formed in any direction within this swath. From just a single ping, it is thus possible to see the highlight of the pipeline, as long as it is within that 10 to 20 degree sector from the AUV's heading.

Additionally, the pipeline and its shadow can robustly be detected from just a single ping, as its straight line response across the sector-scan image distinguishes it from things like pipe diameter-sized rocks that cause spurious detections in single-ping side scan data.

Adaptive Guidance
Once the tracker module identifies the candidate pipeline paths, the paths are sent to the adaptive guidance module, which steers the vehicle along the pipeline. The module then uses the detected track and any prior information to compute the ideal heading for the vehicle in order to optimize the vehicle's view of the pipeline.

Due to the large swath width of the HISAS 1030, there is no need for the vehicle to react to small variations in the detected pipeline track. Instead, the adaptive guidance module can balance the trade-off between keeping the vehicle strictly parallel to the pipeline to maximize the sonar returns and enforcing a well-behaved motion that will provide a stable platform for all of the onboard sensors.

If an LED camera is used for visual inspection during the second half of the run, the pipeline track generated during the first half is used as a reference input to the control system. In this scenario, the AUV follows the generated pipeline trajectory at a low altitude above the pipeline. Camera and/or multibeam data can also be used to update the pipeline trajectory in real time to remove any offsets.

Since the pipeline track is already available when the visual inspection is performed, the AUV can safely survey the pipeline at low altitude and normal survey speed. The known pipeline track is used to plan ahead and generate a smooth trajectory for the AUV that maintains optimal altitude above the pipeline while keeping pitch rates at a minimum.

Long-distance, straight-line missions are the most difficult for autonomous underwater navigation. Even a state-of-the-art inertial navigation system like the one on HUGIN 1000 will have a position error growth of perhaps 10 meters per hour in such a scenario.

However, by comparing the detected pipeline position with its prior-known location, long-term lateral errors can almost be eliminated. If required, along-track errors can be reduced by placing acoustic transponders along the pipeline. Even a single transponder on every 10 to 20 kilometers of pipe will effectively bind the post-processed absolute position error to around five to 10 meters.

After recovery, SAS, camera and multibeam data are downloaded and processed on board the support vessel.

After a 16-hour dive, 30 nautical miles of pipeline will have been surveyed and the following data sets made available: SAS imagery at better than five-by-five-centimeter resolution over a 150-meter corridor centered on the pipeline, SAS bathymetry at 10-to-50-centimeter resolution with the same coverage as the SAS imagery, multibeam bathymetry at 20-to-25-centimeter resolution over a 30-meter corridor centered on the pipeline, and optical images with single-frame coverage of three-by-five to seven-by-10 meters and resolution typically better than one-by-one centimeters (turbidity dependent),with a frame overlap of 50 to 90 percent.

In the alternate scenario of using HISAS on both halves of the survey, SAS imagery and bathymetry from a continuous swath width of 500 to 600 meters is available.

The detection and tracking algorithm can be reapplied to the multisensor post-processed data set to provide an updated reference pipe track. By comparing this to the prior data set, pipeline buckling and changes in burial can be automatically and accurately computed.

The authors are of the opinion that AUV and SAS technology can significantly speed up and improve the quality of pipeline inspection operations. A survey vessel equipped with both an AUV and an ROV can use the AUV for the survey work while the ROV is being used for detailed inspection of detected anomalies.

This is analogous to the development that has taken place in commercial survey operations and military mine countermeasures over the last decade.

The authors would like to thank the staff at Kongsberg Maritime and FFI for providing material and insight to this article. In particular, thanks to Terje Gunnar Fossum at Kongsberg Maritime and Dr. Roy Edgar Hansen and Torstein Olsmo Sæbø of FFI's SAS team.

Per Espen Hagen began his career at the Norwegian Defence Research Establishment in 1990, where he worked on sonar image analysis, nontraditional navigation, operator interfaces and synthetic aperture sonar. In 2008, Hagen joined Kongsberg Maritime, where he is in charge of system architecture for the autonomous underwater vehicle research and development department. He received his M.Sc. in signal processing from the Norwegian Institute of Technology.

Dr. Even Børhaug received his Ph.D. in engineering cybernetics from the Norwegian University of Science and Technology in 2008, with a focus on guidance and control of single and multiple autonomous underwater vehicles (AUVs). He joined Kongsberg Maritime later that year, where he works as a research and development engineer focusing on AUV guidance, control and communications.

In 1996, Øivind Midtgaard joined the autonomous underwater vehicle research group at the Norwegian Defence Research Establishment. His research areas include sonar image processing, automatic target recognition, underwater optical imaging and data fusion. He was previously employed at Robit AS (now Force Technology), where he developed eddy current inspection techniques for offshore flexible pipes. He received his M.Sc. in statistical physics from the Norwegian Institute of Technology.

-back to top-

-back to to Features Index-

Sea Technology is read worldwide in more than 110 countries by management, engineers, scientists and technical personnel working in industry, government and educational research institutions. Readers are involved with oceanographic research, fisheries management, offshore oil and gas exploration and production, undersea defense including antisubmarine warfare, ocean mining and commercial diving.