NRI: Collaborative Research: Enabling Risk-Aware Decision Making in Human-Guided Unmanned Surface Vehicle Teams

NRI：协作研究：在人类引导的无人地面车辆团队中实现风险意识决策

• 批准号: 1634433
• 负责人: Satyandra Gupta
• 金额: $ 52.62万
• 依托单位: University of Southern California
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-01-06 至 2019-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1634433&HistoricalAwards=false
• 关键词: NRI; Collaborative; Research; Enabling; Risk

Over the last ten years, substantial progress has been made in the development of small low-cost unmanned surface vehicles (USVs). There are a number of civilian applications where deploying a human-robot team consisting of several small USVs and one or more human supervisors can significantly reduce costs, improve safety, and increase operational efficiencies. Representative applications include remote/persistent ocean sensing, marine search and rescue, maritime operations in congested port environments, and industrial offshore supply and support. USVs face unique challenges that are not experienced by robots operating indoors, such as: the need to adhere to marine navigation rules (COLREGs); local current, wave and wind conditions that can severely reduce the dynamic range of sensors and actuators; frequent communication interruptions; and risk and urgency due to rapidly changing situations during outdoor on-water operations. This research aims to develop decision making foundations for enabling teams of humans and USVs to perform complex collaborative tasks. Advances in this area could be extremely important from both a regulatory and practical standpoint for the future deployment of USV systems. Results from this research will enable leveraging the tremendous potential of USVs by reducing the cost of deployment and operational risks in civilian applications. The integration of the research with graduate and undergraduate courses will enhance the robotics and ocean engineering curricula and enrich learning experiences of the participating students. Outreach activities will educate and inform K-12 students about career opportunities in marine robotics. The overall goal of the proposed effort is to make advances in risk-informed decision making so that teams of USVs and human supervisors can work cooperatively on a wide variety of missions. The proposed work will develop a comprehensive distributed decision making approach by leveraging the latest advances in task coordination and assignment, planning, reactive behaviors, and control to enable the deployment of human-guided USV teams in civilian applications. Progress in these constituent components will be pursued to ensure that they are consistent with each other and to explicitly account for risk during decision making. This research will develop methodologies to model team missions to ensure that all phases of decision making will have the required information for making informed decisions. Decision making methodologies will be developed for sparse advisory control of USV teams to mitigate risks and for coordinating and assigning tasks to different USVs in the team. Algorithms will also be developed for risk-aware deliberative trajectory planning and generating and executing reactive behaviors for mitigating risks. The methods developed will be validated through on-water field experiments.

在过去的十年里，小型低成本无人水面飞行器(USV)的发展取得了实质性的进展。在许多民用应用中，部署由几个小型无人机和一个或多个人类主管组成的人-机器人团队可以显著降低成本、提高安全性和提高运营效率。具有代表性的应用包括远程/持久海洋遥感、海上搜救、拥挤港口环境中的海上作业，以及工业近海供应和支持。USV面临着在室内操作的机器人没有经历过的独特挑战，例如：需要遵守海洋航行规则(COLREGS)；当地的海流、海浪和风条件可能会严重缩小传感器和执行器的动态范围；频繁的通信中断；以及由于户外水上操作中快速变化的情况而产生的风险和紧迫性。这项研究旨在开发决策基础，使人类和无人机团队能够执行复杂的协作任务。从管理和实践的角度来看，这一领域的进展对未来USV系统的部署可能极其重要。这项研究的结果将能够通过降低民用应用中的部署成本和操作风险来利用USV的巨大潜力。将研究与研究生和本科课程相结合，将加强机器人和海洋工程课程，丰富参与学生的学习经验。外展活动将教育和告知K-12学生在海洋机器人领域的职业机会。拟议工作的总体目标是在了解风险的决策方面取得进展，以便无人机小组和人类监督员能够在各种任务上合作。拟议的工作将开发一种全面的分布式决策方法，通过利用任务协调和分配、规划、反应行为和控制方面的最新进展，使人引导的USV团队能够部署在民用应用中。将继续在这些构成部分方面取得进展，以确保它们彼此一致，并明确说明决策过程中的风险。这项研究将制定方法来模拟小组任务，以确保决策的所有阶段都有作出知情决定所需的信息。将制定决策方法，以便对无人潜航器团队进行稀疏咨询控制，以减轻风险，并协调和分配任务给团队中的不同无人潜航器。还将开发算法，用于风险感知的深思熟虑的轨迹规划，以及生成和执行反应行为，以降低风险。所开发的方法将通过水上现场实验进行验证。