NRI: INT: Adaptive Bio-inspired Co-Robot algorithms for volcano monitoring

NRI：INT：用于火山监测的自适应仿生协作机器人算法

• 批准号: 2024520
• 负责人: Melanie Moses
• 金额: $ 149.54万
• 依托单位: University of New Mexico
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-10-01 至 2024-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2024520&HistoricalAwards=false
• 关键词: NRI; INT; Adaptive; Bio; inspired

There are an estimated 500 volcanoes that emit volcanic gases to the atmosphere. This project will design, build, and field-test a collaborative swarm of flying robots called the Volcano Co-robot Adaptive Natural algorithms (VolCAN) swarm. The VolCAN swarm will transform our ability to forecast volcanic eruptions. The swarm consists of multiple autonomous aerial drones that use algorithms inspired by biology to monitor the unpredictable environments surrounding volcanoes. In addition to monitoring gasses that precede volcanic eruptions, thereby protecting human lives, it will also measure how much carbon dioxide is emitted from volcanoes to better understand how they contribute to the global carbon budget. The VolCAN swarm can adapt to environmental conditions autonomously in real time, and it can also be guided by scientists to collect scientific data during the battery-limited flights of small drones. Our approach leverages the advantages of bio-inspired algorithms that are fast rather than perfectly accurate, and resilient rather than centrally controlled. The project will broaden participation in computing by involving students from underrepresented groups in both robotics research and programming courses.This project will develop, analyze and rigorously test a co-robot swarm of unpiloted air vehicles (UAVs) that collect valuable scientific data in dynamic and unpredictable environments. The VolCAN swarm will use bio-inspired algorithms to detect CO2 plumes, descend plume gradients to measure maximum flux of CO2 from ground sources, estimate plume size, and infer maps of multiple CO2 sources over hundreds of square kilometers. Given battery limitations on flight times and dangerous, unpredictable conditions, the algorithms prioritize speed, robustness and interpretability over high accuracy. The novel bio-inspired algorithms scale to cover vast areas, adapt to the sensed environment to focus monitoring on the most important regions in real time, and are fast enough to collect many simultaneous emissions within the limited battery life of small UAV. Theoretical analyses will determine bounds on the speed and convergence times of the algorithms, and simulations and frequent field tests will measure the performance of the VolCAN system in rigorous, replicated experiments. Additionally, the project will demonstrate that scientists can operate the VolCAN swarm to collect data in the field with either fully autonomous adaptive surveillance or with scientist guidance. This approach combines human flexibility and judgment with the speed and mobility of a UAV swarm. The project will demonstrate the broad applicability of the VolCAN swarm in environmental monitoring applications in experiments to measure methane emissions from pipelines and assess ecological health of plant communities. It will also show that bio-inspired robots can function outside of highly-structured factories, labs, and warehouses to gather valuable scientific data in the hazardous and unpredictable environments of active volcanoes.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.

据估计，有500座火山向大气排放火山气体。这个项目将设计、建造和现场测试一个协作的飞行机器人群，称为火山联合机器人自适应自然算法(VOLCAN)群。火山群将改变我们预测火山喷发的能力。该蜂群由多架自动无人机组成，这些无人机使用受生物学启发的算法来监测火山周围不可预测的环境。除了监测火山喷发前的气体，从而保护人类生命之外，它还将测量火山排放的二氧化碳数量，以更好地了解它们对全球碳预算的贡献。伏尔坎蜂群可以实时自主适应环境条件，还可以在科学家的引导下，在小型无人机电池有限的飞行中收集科学数据。我们的方法利用了受生物启发的算法的优势，这些算法速度快而不是完全准确，具有弹性而不是中央控制。该项目将通过让来自未被充分代表的群体的学生参加机器人学研究和编程课程来扩大对计算的参与。该项目将开发、分析和严格测试无人驾驶飞行器(UAV)的协作机器人群，这些无人机在动态和不可预测的环境中收集有价值的科学数据。伏尔坎蜂群将使用生物启发算法来探测二氧化碳羽流，降低羽流梯度以测量来自地面来源的二氧化碳的最大通量，估计羽流大小，并推断数百平方公里内多个二氧化碳来源的地图。考虑到电池对飞行时间和危险、不可预测条件的限制，这些算法优先考虑速度、稳健性和可解释性，而不是高精度。这种新颖的生物启发算法可以覆盖广阔的区域，适应传感环境，实时监控最重要的区域，并且速度足够快，可以在小型无人机有限的电池寿命内收集许多同时排放的气体。理论分析将决定算法的速度和收敛时间的界限，模拟和频繁的现场测试将在严格的重复实验中测量Volcan系统的性能。此外，该项目将证明，科学家可以通过完全自主的自适应监视或在科学家的指导下操作狼群在野外收集数据。这种方法将人类的灵活性和判断力与无人机群的速度和机动性结合在一起。该项目将展示Volcan群体在环境监测应用中的广泛适用性，这些应用包括测量管道甲烷排放和评估植物群落生态健康的实验。它还将表明，受生物启发的机器人可以在高度结构的工厂、实验室和仓库之外工作，在危险和不可预测的活火山环境中收集有价值的科学数据。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Adaptive Control for Cooperative Aerial Transportation Using Catenary Robots

使用悬链机器人进行协作空中运输的自适应控制

• DOI: 10.1109/airpharo52252.2021.9571068
• 发表时间: 2021
• 期刊: IEEE 2021 Aerial Robotic Systems Physically Interacting with the Environment (AIRPHARO
• 作者: Cardona, Gustavo A.;D'Antonio, Diego S.;Fierro, Rafael;Saldana, David
• 通讯作者: Saldana, David

LoCUS: A Multi-Robot Loss-Tolerant Algorithm for Surveying Volcanic Plumes

• DOI: 10.1109/irc.2020.00025
• 发表时间: 2020-09
• 期刊: 2020 Fourth IEEE International Conference on Robotic Computing (IRC)
• 作者: J. Erickson;Abhinav Aggarwal;G. M. Fricke;M. Moses

Machine learning feature analysis illuminates disparity between E3SM climate models and observed climate change

机器学习特征分析揭示了 E3SM 气候模型与观测到的气候变化之间的差异

• DOI: 10.1016/j.cam.2021.113451
• 发表时间: 2021
• 期刊: Journal of Computational and Applied Mathematics
• 影响因子: 2.4
• 作者: Nichol, J. Jake;Peterson, Matthew G.;Peterson, Kara J.;Fricke, G. Matthew;Moses, Melanie E.
• 通讯作者: Moses, Melanie E.

Boundary Sketching with Asymptotically Optimal Distance and Rotation

具有渐进最佳距离和旋转的边界草图

• 发表时间: 2023
• 期刊: Structural Information and Communication Complexity
• 作者: Dani, Varsha;Islam, Abir;Saia, Jared
• 通讯作者: Saia, Jared

Aerial Survey Robotics in Extreme Environments: Mapping Volcanic CO2 Emissions With Flocking UAVs

极端环境下的航测机器人：利用无人机集群绘制火山二氧化碳排放图

• DOI: 10.3389/fcteg.2022.836720
• 发表时间: 2022
• 期刊: Frontiers in Control Engineering
• 作者: Ericksen, John;Fricke, G. Matthew;Nowicki, Scott;Fischer, Tobias P.;Hayes, Julie C.;Rosenberger, Karissa;Wolf, Samantha R.;Fierro, Rafael;Moses, Melanie E.
• 通讯作者: Moses, Melanie E.