CRII: CPS: Cognitive Trust in Human-Autonomous Vehicle Interactions

这项研究的目标是创造新的技术，以促进我们对信任在人类与自动驾驶汽车互动中的作用的理解，并协助设计未来车辆的安全和值得信赖的自动驾驶。我们正在见证自动驾驶汽车的技术进步。随着车辆自动化程度的提高，人与自动驾驶交互的性质变得更加复杂，需要提出的关键问题是如何确保人与自动驾驶车辆交互的安全性和信任度。一方面，备受瞩目的事件清楚地表明了“过度信任”或过度依赖自动驾驶汽车的风险。另一方面，“信任不足”可能会导致对自动化的忽视或利用不足。这项工作将被整合到研究生和本科生的网络物理系统教育，以及各种科学，技术，工程和数学（STEM）拓展计划的K-12学生。该方法是开发语言和算法，正式表达和推理的信任，在人类自动驾驶汽车的互动。这包括可以代表人类对自动驾驶汽车信任的新模型，以及可以以易于处理的方式使用这些模型来验证信任要求的新形式化方法。主要创新包括基于人类认知和生理状态的多模态感知来测量信任的新方法，表达和验证基于信任的规范的新方法，以及解释人类自主交互中基于信任的决策的新方法。如果成功的话，这项研究将使人与自动驾驶汽车互动中的信任正式化，并使自动驾驶汽车的设计建立适当的信任水平。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This goal of this research is to create new techniques for advancing our understanding of the role of trust within human-autonomous vehicle interactions and assisting in the design of safe and trustworthy autonomy into future vehicles. We are witnessing accelerating technological advances in autonomous vehicles. As the degree of autonomy of vehicles increases and the nature of human-autonomy interactions becomes more complex, key questions that need to be asked are how to ensure safety and trust in human-autonomous vehicle interactions. On the one hand, high-profile incidents make clear the risks from "overtrust" or over-reliance on autonomous vehicles. On the other hand, "undertrust" may cause the neglect or under-utilization of automation. This work will be integrated into the graduate and undergraduate education on cyber-physical systems, as well as various Science, Technology, Engineering and Mathematics (STEM) outreach programs for K-12 students.The approach is to develop languages and algorithms for formally expressing and reasoning about trust in human-autonomous vehicle interactions. This includes new models that can represent a human's trust in autonomous vehicles, together with new formal methods that can use these models in a tractable manner to verify trust requirements. The key innovations include new methods to measure trust based on multi-modal sensing of human cognitive and physiological states, new methods to express and verify trust-based specifications, and new methods to explain trust-based decisions in human-autonomy interactions. If successful, this research will allow the formalization of trust in human-autonomous vehicle interactions and the design of autonomous vehicles that establish appropriate levels of trust.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.