本文为美国麻省理工学院（作者：Hudson D. Graham）的硕士论文，共105页。

随着技术的不断进步，自主无人驾驶飞行器（UAV）的操作很可能由控制多个无人机的单一操作员来执行。由于操作员的注意力分散在多个监督任务上，因此需要支持操作员对任务状态的认识，以实现安全有效的任务管理。

本研究探讨单一操作员对多个无人机控制的音频界面提示与强化改进措施。本文专门评估了连续和离散音频提示的价值，以此作为单无人机和多无人机任务目标航向偏差或延迟到达的指标。特别是，本文提出了两个问题：（1）与离散音频相比，连续音频是否有助于对无人机作战进行人工监控；（2）离散或连续音频支持的有效性是否取决于操作员的工作量？我们在多自主无人机试验台上进行了44名军事参加者的测试。具体来说，两个连续的音频警报被映射到MAUVE内的两个人工监控任务，对这些连续警报进行了单次蜂鸣离散警报测试。研究结果表明，使用连续音频警报可以提高单个操作员监控单个和多个半自主无人机的性能。结果还强调，必须正确地将连续音频与显示器中的其他听觉警报和视觉表示相结合，因为离散音频警报可能被连续音频掩盖，从而使操作员依赖于显示器的视觉提示。

With advanced autonomy, Unmanned AerialVehicle (UAV) operations will likely be conducted by single operatorscontrolling multiple UAVs.    As operator attention is divided acrossmultiple supervisory tasks, there is a need to support the operator’s awarenessof the state of the tasks for safe and effective task management.  This research explores enhancing audio cues of UAV interfaces for thisfuturistic control of multiple UAVs by a single operator. This thesisspecifically assesses the value of continuous and discrete audio cues asindicators of course‐deviations or late‐arrivals to targets for UAV missionswith single and multiple UAVs.    In particular, this thesisaddresses two questions: (1) when compared with discrete audio, does continuousaudio better aid human supervision of UAV operations, and (2) is theeffectiveness of the discrete or continuous audio support dependent on operatorworkload?    An experiment was carried out on the MultipleAutonomous Unmanned Vehicle Experiment (MAUVE) test bed with 44 militaryparticipants. Specifically, two continuous audio alerts were mapped to twohuman supervisory tasks within MAUVE.   These continuous alerts weretested against single beep discrete alerts.    The results show thatthe use of the continuous audio alerts enhances a single operator’s performancein monitoring single and multiple, semi‐autonomous vehicles.  Theresults also emphasize the necessity to properly integrate the continuous audiowith other auditory alarms and visual representations in a display, as it ispossible for discrete audio alerts to be masked by continuous audio, leavingoperators reliant on the visual aspects of the display.   

1 引言

2 项目背景

3 仿真器与接口设计

4 研究方法

5 研究结果

6 讨论

7 结论

附录A 音频警报指南

附录B 情景事件

附录C 参与者同意书

附录D 实验研究的调查统计

附录E MAUVE-MITUS教程

附录F 实验后调查

附录G GLM分析：SPSS输出

更多精彩文章请关注公众号：