RAPID: Collaborative Research: Data Mining and Fusion Between Unmanned Aerial Systems and Social Media Technologies to Improve Emergency Operations

“巴里”飓风、“哈维”飓风等灾害为记录紧急行动中心（eoc）如何通过了解和减轻所涉及的不确定性，利用技术应对自然灾害提供了机会。社交媒体、无人驾驶自主系统/无人机（uav）、实时灾难建模和广泛连接的使用意味着更有效的分析和信息流。有关城市受灾最严重地区的位置或滞留人员的即时信息将挽救生命。实时数据使应急管理部门能够制定更有针对性的反应和灾后恢复计划，这被认为是几十年前尝试的搜索和救援战略在技术上的飞跃。团队拟定的研究计划不仅会收集真实灾害事件的相关信息，还会对数据进行分析和整合，利用无人机等先进技术以及各种数据收集模式，制定灾后管理框架。记录eoc目前的操作效率低下、技术差距和数据分析限制，对于改进灾害准备和响应非常重要。此外，由于在构建框架时需要收集和利用最近一组风暴的时间敏感数据，因此该项目具有时间紧迫性。该RAPID项目将利用“巴里”飓风作为建立框架的机制，该框架将纳入紧急行动中心，以支持灾后分析和决策。飓风巴里伴随着路易斯安那州沿海社区的大面积洪水，这为社交媒体和无人机图像分析提供了一个易腐烂和大量的数据集。该项目将开发用于社交媒体数据挖掘的工具，并与收集的无人机图像融合，用于灾后分析。作为该项目的一部分，EOC运营商和决策者将提供反馈，以增强算法和分析，以支持恢复以及对社交媒体产生的谣言的响应。飓风哈维过后，博蒙特应急中心的数据也将被收集，其中还包括社区恢复的数据。通过结合这两个地区的数据，将产生一个更丰富的数据集，以进行跨空间、时间、灾害程度和社会经济因素的比较分析和联系。RAPID项目最终将收集和归档（在国家科学基金会的自然灾害工程研究基础设施设计安全中）飓风哈维和巴里的EOC操作和技术应用的丰富数据集。然而，至关重要的是，将获得的知识回馈给eoc，以便他们能够为改善美国公民的生活做出改进。研究团队将编写一份手册，记录在Beaumont和Louisiana油田的EOC经验教训，以及如何采用和实施技术。技术的采用和实施需要主动学习来保留知识。因此，将开发为期一天的短期课程，提供无人机和推特演示以及数据分析和融合的示例。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。

Disasters such as Hurricane Barry, Hurricane Harvey and others provide an opportunity to document how emergency operation centers (EOCs) can use technology to respond to a natural disaster by understanding and mitigating the uncertainties involved. The use of social media, Unmanned Autonomous Systems/Unmanned Aerial Vehicles (UAVs), real-time disaster modeling, and widespread connectedness means more efficient analysis and flow of information. Immediate information on the location of most damaged areas of a city or stranded people will save lives. Real-time data allows emergency management to develop more targeted response and post disaster recovery plans, and this is regarded as a technological leap from the previous search and rescue strategies attempted decades ago. The proposed research plan formulated by team will not only collect relevant information from real hazard events, but will also analyze and integrate the data to develop the post disaster management frameworks using advanced technologies including UAVs, and various other modes of data collection. Documenting the current operational inefficiencies, technology gaps, and data analysis limitations of EOCs are important for improvements in disaster preparedness and response. Furthermore, this project has time urgency due to the need to collect and utilize time sensitive data from this recent set of storms in the construction of the framework.This RAPID project will leverage Hurricane Barry as a mechanism for creation of a framework that will be integrated into Emergency Operations Centers (EOCs) to support post-disaster analysis and decision-making. Hurricane Barry was accompanied by extensive flooding in coastal Louisiana communities and this has provided a perishable and voluminous data set of social media and UAV imagery for analysis. The project will develop tools for data mining of the social media and fusion with collected UAV imagery for post-disaster analysis. As part of this project, feedback from EOC operators and decision-makers will be provided that will enable enhancement of algorithms and analyses to support recovery as well as response to social media generated rumors. Data from Beaumont EOC from Hurricane Harvey, which also includes data on the recovery of the community will also be collected. By combining data from both regions, a richer dataset will be produced to make comparative analyses and linkages across space, time, disaster level, and socioeconomic factors. This RAPID project will culminate with collecting and archiving (in the National Science Foundation's Natural Hazards Engineering Research Infrastructure DesignSafe) a rich dataset of EOC operations and technology application from Hurricanes Harvey and Barry. However, it is vital that the knowledge gained be reciprocated back to the EOCs so that they can make improvements for the betterment of US citizens. Research team will develop a manual documenting lessons learned for EOC in Beaumont and Louisiana and how to adopt and implement technology. Technology adoption and implementation requires active learning to retain knowledge. Thus, one-day short courses will be developed that will provide UAV and Twitter demonstrations along with examples of data analysis and fusion.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.