RAPID: Collaborative Research: Unmanned Aerial System Datasets from Hurricanes Harvey and Irma

这个为期12个月的快速项目将整理飓风哈维的119次飞行和飓风伊尔玛的247次飞行的无人机系统(UAS)数据，以便研究人员能够获得这些数据。该项目将(1)筛选个人隐私识别和适用的州法律，并向研究人员提供；(2)主办一个虚拟研讨会，为研究人员介绍数据集和访问程序；(3)在2018年IEEE国际安全与救援机器人研讨会上主办后续研讨会，讨论研究人员使用数据集的结果以及未来如何获取数据集。数据集由德克萨斯州农工机械工程实验站和佛罗里达州立大学收集，他们共同部署到德克萨斯州本德堡县和佛罗里达州普特南和科利尔县。它们之所以独特，是因为它们(1)反映了公职人员的飞行次数最多；(2)代表了在不同高度和八个截然不同的任务飞行的各种UAS平台(10种不同型号)；(3)由基于任务的视频、静止图像和摄影测量数据产品组成；(4)包括活动前、反应阶段和恢复阶段的飞行。然而，这些数据集需要一些额外的工作来筛选隐私内容、完成元数据、保持命名约定的一致性以便于外部研究人员访问以及设置受控访问。这些数据集是短暂的，因为每次任务中丢失的飞行员、平台和飞行时间的元数据将在飞行员的记忆中丢失，而且需要这些数据集来使计算和工程中的其他研究人员使用这些数据。

This 12-month RAPID project will organize unmanned aerial system (UAS) data from 119 flights at Hurricane Harvey and 247 at Hurricane Irma, in order to make the data available to researchers. The project will (1) screen for personal privacy identification and applicable state laws, and make available to researchers, (2) host a virtual workshop to introduce the dataset and access procedures for researchers, and (3) host a follow up workshop at the 2018 IEEE International Symposium on Safety Security and Rescue Robotics to discuss the results by researchers using the datasets and how to acquire datasets in the future.The datasets were collected by the Texas A&M Engineering Experiment Station and Florida State University who co-deployed to Fort Bend County, Texas, and Putnam and Collier Counties, Florida. They are unique because they (1) reflect the largest number of flights for public officials; (2) represent a diversity of UAS platforms (10 different models) flying at different altitudes and for eight distinctly different missions; (3) consist of video, still imagery, and photogrammetric data products based on mission; and (4) include flights before the event, during the response phase, and through the restoration phase. However, the datasets require some additional work to screen content for privacy, to complete meta-data, to be consistent in naming conventions in order to facilitate access by external researchers, and to set up controlled access. The datasets are ephemeral because the missing meta-data of pilots, platforms, and flight times for each mission will be lost in the memory of the pilots and because the datasets are needed to enable other researchers in computing and engineering use the data.