RAPID: Using an Unmanned Aerial Vehicle and Increased Autonomy to Improve an Unmanned Marine Vehicle Lifeguard Assistant Robot

美国面临着大规模海上伤亡事件的风险，如科诗达协和号或世越号渡轮沉没，但幸运的是，EMILY远程操作无人海上水面车辆（USV）等机器人可以作为移动救生员提供帮助。emily现在被希腊海岸警卫队和整个希腊的红十字会使用。救援人员要求EMILY自动导航，这样救生员就可以指定水中的一群人，然后EMILY会自动驾驶到水中的受害者那里，这些受害者会抓住、抓住并等待，而人类救生员则专注于其他由于失去知觉、体温过低的风险（尤其是儿童）或高度痛苦而需要立即援助的人。Robin Murphy教授将与44名研究生和本科生合作，对EMILY进行升级，这样它就可以利用小型系带无人机的视角，结合机载热感测器，自动导航到人群中。该项目将把11名美国学生带到希腊进行两次实地考察，与救生员团队一起积极应对不断变化但仍在进行的乘船难民模式。学生们将参加两个国际研讨会。雅典国立技术大学（NTUA）是这项研究活动的合作伙伴，并将提供热成像专业知识。该项目将向学生介绍与希腊国际合作者的参与性研究，并培训他们在创新的全球工作场所中具有技术竞争力。它有可能拯救希腊和欧洲的生命，并最终为美国的搜救任务做出更有效的贡献。这个为期9个月的项目将有5名研究生和39名本科生参与野外机器人研究。学生将解决如何导航和达到波浪和冲浪的姿势与非完整的，欠驱动的海洋车辆，不能倒车，有动量，和一个有限的转弯半径。他们将采用已知的行为传感器融合方法，利用热图像进行可靠和有效的定位。他们将在不受控制的天气和海浪中从现场的智能EMILY收集数据，并进行额外的适当改进和评估。最后，该项目将产生关于如何与用户合作的纵向案例研究。该研究将通过实施和现场测试有前途的算法和已知原理，推进智能机器人理论——特别是可互操作和可靠的计算机系统、浅海作业中的无人海洋系统、系绳无人机系统、异构团队和人机交互。参与的美国学生将具有全球竞争力，这些数据将为美国公司提供竞争优势，该项目将根据美国和国际法规为自主系统的使用提供服务。该基金由美国国家科学基金会计算机与信息科学与工程理事会（CISE）和国际科学与工程办公室（OISE）资助。

NSF 1637214 RAPID: Using an Unmanned Aerial Vehicle and Increased Autonomy to Improve an Unmanned Marine Vehicle Lifeguard Assistant RobotThe United States is at risk for a mass marine casualty event, such as the Costa Concordia, or MV Sewol ferry sinking, but fortunately robots, such as the EMILY tele-operated unmanned marine surface vehicle (USV) can assist as mobile life preservers. EMILYs are now in use by the Hellenic Coast Guard and Red Cross throughout Greece. The responders have requested that EMILY navigate autonomously so that a lifeguard could designate a cluster of people in the water and then EMILY would autonomously drive to the victims in the water who would grab, hold, and wait while the human lifeguards focused on other people needing immediate assistance due to unconsciousness, risk of hypothermia (especially children), or high distress. Prof. Robin Murphy will engage 44 graduate and undergraduates in implementing an upgrade to EMILY so that it can use the view from a small tethered UAV combined with onboard thermal sensing to autonomously navigate to a cluster of people. The project will bring 11 of these US students to Greece for two field research expeditions to work with lifeguard teams actively responding to the shifting, but ongoing, patterns of boat refugees. The students will participate in two international workshops. The National Technical University of Athens (NTUA) is a partner in this research activity, and will be providing thermal imaging expertise. The project will introduce the students to participatory research with international collaborators in Greece, and train them to be technically competitive in an innovative, global workplace. It will potentially save lives in Greece and Europe, and ultimately contribute more effectively in search-and-rescue missions in the U.S. This 9-month project will involve 5 graduate students and 39 undergraduates in field robotics research. The students will address how to navigate and reach poses in waves and surf with a non-holonomic, under-actuated marine vehicle that cannot back up, has momentum, and a limited turning radius. They will employ known behavioral sensor fusion methods for reliable and efficient navigation targeting people using thermal imagery. They will collect data from the smart EMILY in the field with responders in uncontrolled weather and waves and make additional rounds of appropriate improvements and evaluation. Finally, the project will produce a longitudinal case study of how to work with users. The research will advance the theory of intelligent robotics-- especially interoperable and reliable computer systems, unmanned marine systems in shallow littoral operations, tethered unmanned aerial systems, heterogeneous teams, and human-robot interaction-- by implementing and field testing promising algorithms and known principles. The participating American students will be globally competitive, the data will give US companies a competitive edge, and the project will serve the use of autonomous systems in accordance with US and international regulations.This grant was funded by NSF Directorate of Computer & Information Science & Engineering (CISE), and Office of International Science and Engineering (OISE).