The inspection of subsea pipelines is carried out under tight cost and performance constraints. Over recent years, the technology has emerged to carry out such inspection primarily using sidescan sonar. Current approached to the interpretation of sidescan sonar data rely very heavily on human interpretation. This paper presents a computer-based approach to the interpretation of sidescan pipeline survey data, firstly by improving the productivity of human operators, and then by automating portions of the inspection process itself.
There are over 12,500 kilometers of pipelines in the North Sea. Although legislation has resulted in the burial of recently laid pipelines, thousands of kilometers of pipeline have been laid directly on the seabed or in shallow trenches. Operators are required to carry out frequent inspections of these pipelines, either visually or by using sidescan sonar. As a means of decreasing inspection costs, sidescan surveys have become increasingly attractive, having the potential to be carried out an order of magnitude faster than visual (ROV) surveys. The unremitting downwards pressure on costs-fuelled by intense competition for survey contracts-has focused attention on the sidescan survey process-in particular, the reporting timescales and efficient use of offshore manpower.
The mechanics of sidescan pipeline inspection are fairly well established thermal paper records are obtained in an "on-line" survey suite, and are conveyed to the "off-line" room for inspection and digitisation by geophysicists. The information from the paper traces is accompanied by the sidescan imagery of the area.
Schedules for preliminary reports tend to be tight-often they must be submitted between 24 and 48 hours after the vessel reaches port, therefore this stage of reporting must be carried out offshore. The detailed report tends to be produced on a more leisurely timescale, but controlling these timescales and thereby the costs of detailed reporting has proved problematical.
Two areas in the sidescan survey process seem to be candidates for improvement
the use of the time of offshore personnel - geophysicists and surveyors-could be made more efficient
as much as possible of the reporting process should be completed offshore, minimizing the amount of onshore time
Clearly, if these two objectives could be perfectly combined, the result would be that the offshore survey crew could spend 100% of their time working towards producing a final report which would be available on arrival onshore.
This objective is the motivation of Coda Technologies1 approach to automisation of the sidescan sonar pipeline survey process. The point, as with any technological changes, is not change for change's sake, or because the technology is exciting or in some unquantifiable way "good", rather, new technology should offer concrete benefits in the shape either of increased efficiency or decreased costs. When equipment and procedures are relatively mature-as is the case for sidescan sonar surveying - there is little scope for piecemeal changes and improvements. Benefits are most likely to flow from a holistic approach-an examination of every stage of the process from putting the towfish in the water
