RAPID: Human-Robotic Interactions During Harvey Recovery Operations

有效和高效的灾难恢复是个人、社区和企业从飓风哈维和伊尔玛等大规模灾难中恢复正常运作所必需的。近年来，无人驾驶机器人被用于促进救援、反应和恢复，并在这些努力中被发现是无价的，因为它们可以到达人类无法到达的地方。虽然这些机器人本身没有人在车上，但它们确实需要人类来操作，而且在灾难恢复期间，人们对这种技术工作环境对人类的需求知之甚少。已知的是，飞行员经常工作：在工作要求高、压力大的环境中；在室外(通常是在炎热的夏天，因为飓风会发生)；需要移动或长时间以尴尬的姿势站立；有时这些操作员自己住在受影响的地区，因此可能会因为灾难而经历心理和社会压力。鉴于训练有素的操作员数量有限，不同类型的机器人可用，而且该国越来越多的地区需要评估因大规模灾害而使用机器人造成的损害，因此迫切需要审查受影响地区恢复行动期间人与机器人之间的自然互动。这项研究将对在灾难恢复操作中具有弹性的人类和机器人团队之间的协作属性建立一个基本的、原则性的理解，以最大限度地减少代价高昂的错误，并提高未来灾难机器人响应和恢复操作的有效性。这项快速奖励将提供关于在德克萨斯州墨西哥湾沿岸和受洪水影响的周围地点进行机器人辅助的哈维恢复操作期间人/机器人交互的关键和及时的信息。这项研究将审查恢复行动，重点是检查受洪水影响的关键基础设施，并帮助不同类型的建筑物(房屋、工厂、社区等)的经济复苏。这项研究的目的是1)记录人(如飞行员)和机器人(如无人飞行器)之间的关系，以在动态变化和不稳定的环境(如洪水破坏的基础设施)中完成特定的恢复任务(监视和检查)；2)确定人/机器人交互不良的关键因素，为改进人/机器人交互提供启发式/指导方针。将使用定性和定量数据收集和分析技术：视频观察，以记录恢复操作期间人/机器人交互的范围、对工作量/疲劳的感知、对机器人的信任、可用性、沟通和培训差距，通过对人类团队的调查和访谈，所使用的机器人的类型和功能，操作员的生理反应，以及任务生产率指标。从这项研究中获得的结论将迅速传播给适当的利益攸关方(行业、政府、公共安全)，以便在灾后恢复行动中制定有效的最佳做法。

Effective and efficient disaster recovery is necessary for individuals, the community and businesses to return to normal functioning from large-scale disasters, like hurricanes Harvey and Irma. In recent years, unmanned robots have been used to facilitate rescue, response, and recovery and have been found invaluable in these efforts as they can go where humans cannot. Although these robots do not have someone on the vehicle itself, they do require humans to operate them, and little is known about the demands of this technological work environment on the humans during disaster recovery. What is known is that the pilots often work: in high work demand stressful environments; outside (often in the heat, as hurricanes happen in the summer); require ambulation or prolonged standing in awkward postures for extended periods of time; and sometimes these operators live in the affected area themselves and thus maybe experiencing psycho-social stressors due to the disaster. Given the finite number of trained operators, the availability of different types of robots, and the increasing areas of the country needing assessment of damages using robots due to large-scale disasters, there is a critical need to examine naturalistic human/robotic interactions during recovery operations in affected regions. The study will create a fundamental, principled understanding of attributes of collaborations between human and robot teams that are resilient during disaster recovery operations to minimize costly errors and improve effectiveness of future disaster robotics response and recovery operations.This RAPID award will provide critical and timely information on human/robotic interactions during robot-assisted Harvey recovery operations in the Texas Gulf Coast and surrounding locations impacted by flooding. The study will examine recovery operations that focus on inspections of critical infrastructure affected by the flooding and to assist with economic recovery across different types of structures (homes, factories, neighborhoods, etc.). The objectives of this study are to 1) document the relationships between the human (e.g., pilot) and the robot (e.g., unmanned aerial vehicle) to achieve specific recovery tasks (surveillance and inspections) in dynamically changing and unstable environments (e.g., flood-damaged infrastructure); and 2) determine the key contributors of poor human/robotic interactions to provide heuristics/guidelines for improved human/robotic interactions. Both qualitative and quantitative data collection and analyses techniques will be used: video observations to document the gamut of human/robot interactions during recovery operations, perceptions of workload/fatigue, trust in robots, usability, communication, and training gaps through surveys and interviews from the human teams, types and functions of robots used, operator physiological responses, and task productivity metrics. Findings obtained from this study will be rapidly disseminated to appropriate stakeholders (industry, government, public safety) for developing effective best practices in disaster recovery operations.