Autonomous Wind-driven Devices for Acoustic and Visual Sensing

Oshen设计了用于海洋传感的风力驱动、自主水面舰船。自主控制和使用风力发电使我们的设备能够收集精确的数据，这比现有的解决方案更具成本效益。我们的设计只有1.2米长，就像一艘微型帆船。较小的体积使单位成本较低，并允许快速整体部署设备，为持续监测海洋生物多样性提供传感器网络。在一个日益依赖海洋数据的世界里，这种整体技术有助于解决数据稀疏的问题。该项目旨在调查在沿海水域使用这些设备进行两种应用的可行性：声学测量和视觉数据集采集。声学调查可行性的主要重点将是海洋哺乳动物的探测，而视觉数据集收集的主要重点将是海鸟的探测和识别。该项目将分为三个主要阶段：第一，传感器与我们现有设备的物理硬件集成，包括电子系统集成(确保传感器工作并链接回定制的人工智能软件)。第二阶段是一项小型测试，其中一种设备部署在东海岸外的一个地区：例如斯卡伯勒，那里的海豚等海洋哺乳动物很常见，但海上交通密度较低。最后一次规模更大、持续时间更长的测试将使用多达10个设备，覆盖东海岸更广泛的地区。这一地区是来自弗兰伯勒海德和本普顿悬崖特别保护区(SPA)的繁殖海鸟的觅食范围，特别是北甘露-桑树。在海洋哺乳动物中，港口江豚和小须鲸是最有可能在该地区发现的种群。这项测试产生的数据将被处理并用于验证设备识别海洋生物的能力，并检查收集视觉数据集的可行性。

Oshen designs wind-driven, autonomous surface vessels for marine sensing. The autonomous control and use of wind power allow our devices to gather precise data that is significantly more cost-effective than existing solutions. Just 1.2m in length, our design resembles a miniature sailboat. The small size keeps the unit cost low and allows devices to be quickly deployed en-masse, providing a sensor network for continuous monitoring of marine biodiversity. This holistic technology helps solve the problem of data sparsity in a world that is becoming increasingly reliant on ocean data.This project aims to investigate the feasibility of using these devices in coastal waters for two applications: acoustic surveying and visual dataset collection. The primary focus of the acoustic survey feasibility will be marine mammal detection and the primary focus of the visual dataset collection will be seabird detection and identification. There will be three overarching stages to the project: first, the physical hardware integration of the sensors into our existing devices, including the electronic systems integration (ensuring the sensors work and link back to bespoke AI software). The second stage is a small test, where one device is deployed in an area off the East coast: for example, Scarborough, where marine mammals such as harbour porpoises are common but marine traffic has a lower density. The final larger scale, longer duration test will be conducted with up to ten devices covering a broader area along the East coast. In this area lies the foraging range for breeding seabirds from Flamborough Head & Bempton Cliffs Special Protection Area (SPA), notably northern gannet _Morus bassanus_. Among marine mammals, harbour porpoise and minke whales are the populations most likely to be identified in this area. The data resulting from this test will be processed and used to validate the ability of the devices to identify marine life and to examine the feasibility of collecting visual datasets.