NeTS: Large: Collaborative Research: ASTRO: A Platform for 3-D Data-Driven Mobile Sensing via Networked Drones

NeTS：大型：协作研究：ASTRO：通过联网无人机进行 3D 数据驱动的移动传感平台

基本信息

批准号：
1801865
负责人：
Edward Knightly
金额：
$ 150万
依托单位：
William Marsh Rice University
依托单位国家：
美国
项目类别：
Continuing Grant
财政年份：
2018
资助国家：
美国
起止时间：
2018-08-15 至 2023-07-31
项目状态：
已结题

来源：
https://www.nsf.gov/awardsearch/showAward?AWD_ID=1801865&HistoricalAwards=false
关键词：
NeTS Large Collaborative Research ASTRO

项目摘要

The driving vision of this project is to detect Volatile Organic Compounds (VOCs) through ASTRO, a platform for autonomous 3-D data-driven mobile sensing via networked drones equipped with gas sensors. VOCs are hazardous to human health and the environment; they are released by explosions, gas leaks, and industrial accidents prevalent in low-income and under-resourced urban neighborhoods in close proximity to industrial processing plants, chemical refineries, and other sources of airborne pollutants. The project is located in an economically disadvantaged area of Houston, Texas. With Technology For All (TFA), the project team has a history of engaging the local community via broadband access, technology training, and connected health. The TFA wireless network already serves 1000's of community members in several square kilometers in Houston's East End via a mix of commercial Wi-Fi and software defined radios. The project targets realizing a high-resolution ground truth of environmental conditions in low-income urban areas which can impact emergency response procedures and environmental justice via policy and law. The project will develop a mobile app that alerts community residents of hazardous VOC concentrations near their current location. This project will impact urban areas with a demonstration of fusing next generation environmental sensing with next generation wireless access via networked drones. The project's objective is to realize an unprecedented resolution in VOC sensing by development and demonstration of ASTRO, a system for networked drone sensing missions without ground control. ASTRO will realize the unique capability to dynamically move sensors in 3-D according to real-time measurements. Consequently, networks of drones with on-board sensors can find and track VOC plumes, solely by coordinating among themselves, and without requiring a centralized ground controller. Two inter-related thrusts will realize this vision. The first is target detection, tracking, and modeling high VOC concentration clusters, targeting health and environmental safety. The second is development of the underlying principles and methodologies for data-driven mobile missions via drone networks. The project's outcomes will include lightweight machine learning methods that provide foundations for real-time distributed autonomous sensing with environmental and health objectives. These data sets will yield development of atmospheric models of VOCs at a finer resolution than is possible today. Moreover, the outcomes will also include methods for adaptive communication among the networked drones via software defined radios that can adapt their network topology and spectrum usage to realize mission objectives.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.

该项目的主要愿景是通过 ASTRO 检测挥发性有机化合物 (VOC)，ASTRO 是一个通过配备气体传感器的联网无人机进行自主 3D 数据驱动移动传感的平台。 VOCs对人体健康和环境有害；它们是由靠近工业加工厂、炼油厂和其他空气污染物源的低收入和资源贫乏城市社区普遍存在的爆炸、煤气泄漏和工业事故释放的。该项目位于德克萨斯州休斯顿经济贫困地区。通过“全民技术”(TFA)，项目团队一直通过宽带接入、技术培训和互联健康与当地社区互动。 TFA 无线网络已通过商业 Wi-Fi 和软件定义无线电的组合为休斯顿东区数平方公里的数千名社区成员提供服务。该项目的目标是实现低收入城市地区环境条件的高分辨率地面实况，这可以通过政策和法律影响应急响应程序和环境正义。该项目将开发一款移动应用程序，提醒社区居民当前位置附近的有害挥发性有机化合物浓度。该项目将通过展示通过联网无人机将下一代环境传感与下一代无线接入相融合来影响城市地区。该项目的目标是通过开发和演示 ASTRO（一种无需地面控制的网络无人机传感任务系统）来实现 VOC 传感前所未有的分辨率。 ASTRO 将实现根据实时测量动态移动 3D 传感器的独特功能。因此，带有机载传感器的无人机网络可以仅通过它们之间的协调来发现和跟踪 VOC 羽流，而不需要集中的地面控制器。两个相互关联的推动力将实现这一愿景。第一个是目标检测、跟踪和建模高 VOC 浓度簇，以健康和环境安全为目标。第二个是开发通过无人机网络进行数据驱动的移动任务的基本原理和方法。该项目的成果将包括轻量级机器学习方法，为具有环境和健康目标的实时分布式自主传感提供基础。这些数据集将以比目前更精细的分辨率开发 VOC 大气模型。此外，成果还将包括通过软件定义无线电在联网无人机之间进行自适应通信的方法，这些方法可以调整其网络拓扑和频谱使用以实现任务目标。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。

项目成果

期刊论文数量（20）

专著数量（0）

科研奖励数量（0）

会议论文数量（0）

专利数量（0）

PipeGCN: Efficient Full-Graph Training of Graph Convolutional Networks with Pipelined Feature Communication

PipeGCN：具有流水线特征通信的图卷积网络的高效全图训练

DOI：
10.48550/arxiv.2203.10428
发表时间：
2022-03-20
期刊：
ArXiv
影响因子：
0
作者：
Cheng Wan;Youjie Li;Cameron R. Wolfe;Anastasios Kyrillidis;Namjae Kim;Yingyan Lin
通讯作者：
Yingyan Lin

CPT: Efficient Deep Neural Network Training via Cyclic Precision

CPT：通过循环精度进行高效深度神经网络训练

DOI：
发表时间：
2021-01-25
期刊：
ArXiv
影响因子：
0
作者：
Y. Fu;Han Guo;Meng Li;Xin Yang;Yining Ding;V. Ch;ra;ra;Yingyan Lin
通讯作者：
Yingyan Lin

DepthShrinker: A New Compression Paradigm Towards Boosting Real-Hardware Efficiency of Compact Neural Networks

DepthShrinker：一种新的压缩范式，旨在提高紧凑型神经网络的实际硬件效率

DOI：
10.48550/arxiv.2206.00843
发表时间：
2022-06-02
期刊：
ArXiv
影响因子：
0
作者：
Y. Fu;Haichuan Yang;Jiayi Yuan;Meng Li;Cheng Wan;Raghuraman Krishnamoorthi;Vikas Ch;ra;ra;Yingyan Lin
通讯作者：
Yingyan Lin

Quartz-enhanced photoacoustic sensor for ethylene detection implementing optimized custom tuning fork-based spectrophone

用于乙烯检测的石英增强光声传感器实施优化的定制音叉分光器

DOI：
10.1364/oe.27.004271
发表时间：
2019-01
期刊：
Optics Express
影响因子：
3.8
作者：
Giglio, Marilena;Elefante, Arianna;Patimisco, Pietro;Sampaolo, Angelo;Sgobba, Fabrizio;Rossmadl, Hubert;Mackowiak, Verena;Wu, Hongpeng;Tittel, Frank K.;Dong, Lei;et al
通讯作者：
et al

Robust Environmental Sensing Using UAVs

使用无人机进行强大的环境传感