Abstract
DeepOcean and partners have for multiple years been developing an Autonomous Inspection Drone (AID). This is a remotely operated vehicle designed to perform high-quality inspections of subsea assets. The AID is equipped with a state-of-the-art machine vision camera, which is used to autonomously navigate through a pre-programmed inspection scope. The operation of the AID is supervised by our remote operation center in Haugesund, Norway. This paper presents results from our first field trial.
A digital twin was constructed for multiple subsea templates. The digital twin was developed into a game where a model of the AID was controlled. This allowed inspection engineers to plan the route for the inspection. The planned route in the digital twin was exported to the AID. The AID autonomously executed the planned route. Autonomous navigation was made possible by combining an inertial navigation system, a USBL positioning system and a machine vision camera. The machine vision camera detected QR-codes on the subsea asset. This allowed the AID to correct position and determine the true location on the asset.
Multiple tests were performed during this field trial. Observations of these tests are presented in the paper. Examples of tests that were performed during the field trials include autonomous undocking and docking from TMS, executing of pre-programmed inspection route and station keeping when QR codes were detected on the subsea asset. It was observed in the field trials that the control system to the AID would need to be further developed. This is to increase the robustness and enhance the efficiency of autonomous navigation. These considerations are discussed in the paper, and conclusions are drawn with regard to how this may be achieved.
This paper will explain how DeepOcean planned and executed an inspection scope by combining a digital twin of the asset and a machine vision camera used to assist the navigation of the AID. The results and conclusions presented from this field trial are essential for the petroleum industry to step into a future where resident vehicles are able to perform value-creating work autonomously. In the near term, this work is forming the foundation for DeepOcean to increase the performance of simultaneous operations from our vessels. This will be made possible as the AID may be launched from the vessel, which is completing some other intervention work, and come back after the autonomous inspection has been completed. The operation may be supervised by our control centre in Haugesund, Norway
