Collaborative Research: Ideas Lab: Smarter Microbial Observatories for Realtime ExperimentS (SMORES)

大部分海底覆盖着沉积物，其中蕴藏着大量的动物、微生物、细菌和古细菌。这些群落在自然碳循环中发挥着作用，但关于这些群落如何滋养海洋并将碳储存到更深的沉积物中，还有很多东西需要了解。这个创意实验室：推进水下科学的工程技术 (ETAUS) 项目将开发一种新的海底传感器/采样器系统，以研究潮汐如何影响海底碳循环。它还将开发“更智能”的控制系统，以预测何时何地最好地收集样本。这将是对浪费电池电量且部署时间短的现有方法的重大改进。这些进步还将提供可应用于其他传感器组的技术，从而使它们能够更长时间地记录数据，从而产生更广泛的影响。该项目为来自不同背景的高中生和大学生提供指导和实地考察的机会。该团队将与一位获奖电影制作人合作制作一部有关海洋科学职业的短片。该项目将开发一种新型海底传感器/采样器阵列，以更好地了解潮汐抽水和地下洋流如何影响海底氧合和沉积碳循环。同样重要的是，它将开发使用机器学习模型（MLM）的控制系统，根据历史和实时数据对何时何地进行最佳感知和采样做出明智的预测。 MLM控制器的开发将增强远程部署、最大限度地减少多余的功耗并减少次优样本的收集。该项目由以下 NSF 部门联合资助：电气、通信和网络系统、环境生物学、海洋科学以及土木、机械和制造创新。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Much of the ocean floor is covered with sediments that harbor large communities of animals, microscopic creatures, bacteria and archaea. These communities play a role in natural carbon cycling, yet there is a lot to learn about how this community nourishes the ocean, and stores carbon into deeper sediments. This Ideas Lab: Engineering Technologies to Advance Underwater Sciences (ETAUS) project will develop a new seafloor sensor/sampler system to study how tides affect the seafloor carbon cycle. It will also develop “smarter” control systems to make predictions about when and where to best collect samples. This would be a major improvement over existing approaches that waste battery power and have short deployment times. These advances will also provide broader impacts by providing technologies that can be applied to other sensor sets allowing them to record data for longer periods. The project provides mentoring and field opportunities to high-school and college students from a variety of backgrounds. The team will work with an award-winning filmmaker to produce a short documentary about careers in marine science.This project will develop a novel seafloor sensor/sampler array to better understand how tidal pumping and subsurface currents influence seafloor oxygenation and sedimentary carbon cycling. Equally important, it will develop control systems that use machine-learning models (MLM) to make informed predictions about when and where to best sense and sample based on historical and realtime data. The development of MLM controllers will enhance remote deployment, minimize excess power consumption and decrease the collection of sub-optimal samples. This project is funded jointly by the following NSF Divisions: Electrical, Communications & Cyber Systems, Environmental Biology, Ocean Sciences, and Civil, Mechanical & Manufacturing Innovation.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.