Abstract
This paper outlines the remote visual underwater inspection technology using the Lantern Eye Stereo Remote system, a non-contact stereo imaging system that can be deployed without an imaging engineer offshore, allowing scaled 3D models and measurements. This system allows an imaging engineer to operate remotely, linking the mounted camera system used off-shore on an ROV, to an on-shore office. The connection was facilitated by a satellite connection to the ROV Support Vessel (RSV).
The approach was achieved through a novel communication architecture using a deployed computing system, VPN and multiple subnets. The communications are enabled through a robotics middleware that handles the publishing and subscribing of information including compressed images, and commands that change the exposure and gain of the camera system. Additionally, real-time photogrammetric reconstruction allows the on-shore user to understand in real-time the coverage that has been achieved on the mapped object. This makes sure that the required data has been captured, as the data collection process can be complex and tracking the collected data is error prone.
The system allowed multiple successful deployments, without the requirement for the imagine engineer to be offshore with the camera system during data collection. Results show that the system can be applied to any offshore deployment. In situations where the bandwidth is reduced, there is scope for increased compression on the images, but there exists hard requirements on bandwidth and latency where performance degrades beyond feasible usage, which has been determined to be 0.6MB/second (5Mb/s), 1500 ms latency. In the case of stable water conditions, a higher latency can be tolerated due to the simpler data capture process. For future work, greater reliance on the computing platform that is deployed with the camera system is possible, which includes automation of the role of the onshore imaging engineer, including automatic adaptation of image parameters without human intervention, and ensuring coverage is attained without a human monitoring this live. This opens the scope of future work where greater automation is undertaken with significantly less human intervention, for situations with low bandwidth and high latency.
This paper adds to the state of knowledge which benefits the oil and gas industry through showing an operating example of the remotely enabled subsea integrity assessment. This reduces the personnel required offshore and opens the path to further automation of the data collection process, especially in cases of reduced bandwidth or high latency. This opens up further benefits such as less vessel time.
