Autonomous Techniques for anthropogenic Structure Ecological Assessment (AT-SEA)

人为结构生态评估自主技术（AT-SEA）

• 批准号: NE/T010592/1
• 负责人: Blair Thornton
• 金额: $ 21.3万
• 依托单位: University of Southampton
• 依托单位国家: 英国
• 项目类别: Research Grant
• 财政年份: 2021
• 资助国家: 英国
• 起止时间: 2021 至 无数据
• 项目状态: 未结题
• 来源: https://gtr.ukri.org/projects?ref=NE%2FT010592%2F1
• 关键词: Autonomous; Techniques; anthropogenic; Structure; Ecological

Thousands of Oil & Gas industry structures in the sea are approaching the end of their lives. At this time, they typically need to be removed and the environment returned to a safe state. This process is known as decommissioning. As many of these sites are old (typically 20+ years) and originally were drilled before the current environmental regulations existed, there has often been some contamination of the seabed around these sites. To ensure that no harmful effects will occur, decommissioning operations need to be supported by an environmental assessment and subsequent monitoring. Monitoring may be required over many years after decommissioning, especially if some structures are left in place. Monitoring surveys in the offshore environment are expensive and time-consuming, requiring ships and many specialist seagoing personnel. This requirement, although vital, will have a considerable cost for industry and the public.Ocean robots, which use computer systems to carry out survey missions by themselves, are regularly used in detailed scientific assessments of the environment. As they collect very high-quality data quickly, such robots have recently been adopted for some tasks by industry but these still require an expensive support ship as they are not capable of long-range missions. Recent technological developments have cut the cost and expanded the range of these robots to thousands of kilometres, making it possible for long-range assessments of multiple sites to be undertaken with a robot launched from the shore. This would have many advantages, improving the quality and quantity of environmental information while cutting the costly requirement for a survey ship and crew. We will carry out the first fully autonomous environmental assessment of multiple decommissioning sites. The Autosub long-range ocean robot submarine ("Boaty McBoatface") will be launched from the shore in Shetland, visit and carry out an environmental assessment at three decommissioning sites in the northern North Sea, before returning around 10 days later with the detailed survey information onboard. The robot will take photographs of the seabed, and these will be automatically stitched together to make a map of the seafloor, structures present, and the animals that live there. Established sensor systems will measure a range of properties of the water, including the presence of oil and gas. As well as the decommissioned sites, the robot will visit a special marine protected area where we know there are natural leaks of gas, to check the robot can reliably detect a leak if it did occur.On return to shore, the project will examine all the data obtained and compare it to that gathered using standard survey ship methods. We will test if the same environmental trends can be identified from both datasets to determine if the automated approach would be a suitable replacement for standard survey ship operations. The project will also produce a fully documented case study, which includes detailed information on the costs and benefits, practical information on deployments and approaches to reduce the risks and improve the efficiency of operations. This will be used by industry, scientists and government regulators, to demonstrate the techniques and will provide the necessary information to potential users to aid in their adoption. The overall goal of the project is to improve the environmental protection of the North Sea at a reduced cost and to demonstrate how this leading UK robotic technology could be used worldwide.

海洋中成千上万的石油行业结构正在临近他们的生命。此时，通常需要将它们删除，并且环境恢复到了安全状态。这个过程称为退役。由于这些地点中的许多地方都是旧的（通常为20年以上），并且最初是在当前环境法规存在之前进行的，因此通常在这些地点周围的海床污染了一些污染。为了确保不会发生有害影响，需要通过环境评估和随后的监控来支持退役操作。退役后多年可能需要进行监控，尤其是在留下某些结构的情况下。在离岸环境中的监视调查是昂贵且耗时的，需要船只和许多专业的海上人员。这项要求虽然至关重要，但对于行业和公共机器人来说，使用计算机系统自己执行调查任务的欧洲机器人将定期用于环境的详细科学评估。当他们迅速收集非常高质量的数据时，这些机器人最近被某些任务采用了，但是这些机器人仍然需要昂贵的支持船，因为它们无法执行远程任务。最近的技术发展已将这些机器人的成本降低，并将这些机器人的范围扩大到数千公里，从而可以通过从海岸推出的机器人进行多个地点进行长期评估。这将具有许多优势，可以提高环境信息的质量和数量，同时减少对调查船和机组人员的昂贵要求。我们将对多个退役站点进行第一个完全自主的环境评估。 AutoSub远程海洋机器人潜艇（“ Boaty McBoatface”）将从设得兰群岛的海岸发射，在北海北部的三个退役地点进行访问并进行环境评估，然后大约10天后返回，并在船上提供了详细的调查信息。机器人将拍摄海床的照片，这些照片将自动缝合在一起，以制作海底，建筑物以及居住在那里的动物的地图。已建立的传感器系统将测量水的一系列特性，包括石油和天然气的存在。除退役地点外，机器人还将访问一个特殊的海洋保护区，我们知道有天然气体的自然泄漏，以检查机器人是否确实发生了泄漏。如果确实发生了泄漏。返回岸上，该项目将检查所有获得的数据，并将其与使用标准调查船一起收集的数据进行比较。我们将测试是否可以从两个数据集中确定相同的环境趋势，以确定自动化方法是否是标准调查船舶运营的合适替代方法。该项目还将产生完整的案例研究，其中包括有关成本和收益的详细信息，有关部署的实用信息以及降低风险并提高操作效率的方法。行业，科学家和政府监管机构将使用此技术来证明这些技术，并将为潜在用户提供必要的信息，以帮助他们采用。该项目的总体目标是以降低的成本来改善北海的环境保护，并证明如何在全球使用这项领先的英国机器人技术。

项目成果

High-resolution visual seafloor mapping and classification using long range capable AUV for ship-free benthic surveys

使用远程 AUV 进行高分辨率视觉海底测绘和分类，以进行无船海底调查

• 发表时间: 2023
• 作者: Bodenmann A

GeoCLR: Georeference Contrastive Learning for Efficient Seafloor Image Interpretation

• DOI: 10.55417/fr.2022037
• 发表时间: 2021-08
• 期刊: ArXiv
• 作者: Takaki Yamada;A. Prügel-Bennett;Stefan B. Williams;O. Pizarro;B. Thornton

Auto-calibration of line-laser structured-light seafloor mapping systems

线激光结构光海底测绘系统的自动校准

• DOI: 10.23919/oceans44145.2021.9705873
• 发表时间: 2021
• 作者: Stanley D

Leveraging Metadata in Representation Learning With Georeferenced Seafloor Imagery

利用地理参考海底图像的表示学习中的元数据

• DOI: 10.1109/lra.2021.3101881
• 发表时间: 2021
• 期刊: IEEE Robotics and Automation Letters
• 影响因子: 5.2
• 作者: Yamada T