Collaborative Research: EAGER: The next crisis for coral reefs is how to study vanishing coral species; AUVs equipped with AI may be the only tool for the job

珊瑚礁危机的特点是全球珊瑚和鱼类的丰度下降，而大型藻类的丰度上升。该项目利用人工智能（AI）驱动的水下机器人在自主部署中评估了两种稀有珊瑚物种的丰度和灭绝风险。这项新技术可以对珊瑚进行量化，这些珊瑚的种群密度非常低，以至于人类潜水员无法在水下花费足够长的时间来找到它们，这项技术也适用于其他研究领域，改变了对稀有动物的研究。该项目还在西班牙服务的四年制大学和海洋研究机构之间建立了学术桥梁，为两名研究生和一名早期职业研究工程师提供珊瑚生态学和水下机器人的交叉培训，并为当地高中生提供机器人主题培训。珊瑚礁的珊瑚已经枯竭，许多物种正在进入一段罕见的时期，这可能是灭绝的前奏。传统的生态工具不足以列举这些珊瑚，这是量化灭绝风险和恢复前景的第一步。该项目训练一个具有人工智能（AI）的自主水下航行器（AUV）来搜索稀有珊瑚，并在一个由37个珊瑚礁数据创建的试验台中挑战它，这些珊瑚礁的标志性支柱和脑珊瑚正在濒临灭绝。该项目团队正在测试两个假设：(1)最近的事件催化了珊瑚数量的新下降，(2)稀有珊瑚的人口统计轨迹将以灭绝告终。这项新技术适用于其他生态系统，改变了对稀有生物的研究。该项目由地球科学理事会共同资助，旨在支持人工智能/机器学习在地球科学领域的发展。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

The coral reef crisis is characterized by global declines in the abundance of corals and fishes, and a rise in the abundance of macroalgae. This project assesses the abundance and extinction risk of two rare coral species using artificial intelligence (AI)-powered underwater robots in autonomous deployments. This novel technology allows quantification of corals that occur at such low population densities that human divers cannot spend long enough underwater to find them, and is applicable to other fields of research, transforming the study of rare animals. The project also builds an academic bridge between a Hispanic Serving 4-year university and a marine research institution, provides cross-training for two graduate students and an early-career research engineer in coral ecology and underwater robotics, and supports robotics-themed training for local high school students. Coral reefs have become depleted of corals, and many species are entering a period of rarity that might be a prelude to extinction. Traditional ecological tools are inadequate to enumerate these corals, which is the first step in quantifying the risk of extinction and the prospects for recovery. This project trains an autonomous underwater vehicle (AUV) with artificial intelligence (AI) to search for rare corals and challenges it in a test bed created by 37 y of data from coral reefs where iconic pillar and brain corals are being pushed to extinction. The project team is testing two hypotheses: (1) recent events have catalyzed a new decline in abundance of corals, and (2) demographic trajectories of rare corals will end in extinction. This novel technology is applicable to other ecosystems, transforming the study of rare organisms. This project is co-funded by the Directorate for Geosciences to support AI/ML advancement in the geosciences.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.