CPS: TTP Option: Medium: Identifying, Characterizing, and Shaping Multi-Scale Cyber-Human Interactions in Mixed Autonomous/Conventional Vehicle Traffic

智能交通系统的前景是安全性、移动性和效率远远超过目前的技术。然而，交通工程师和规划者必须预测未来的交通需求，使技术迅速发展，而从旧的到新的交通技术的转换将需要几十年的时间。该项目旨在了解这些趋势对交通工程的影响，为自动驾驶车辆和驾驶员之间的有效共享控制提供方法，并为实现智能交通系统的承诺提供战略见解。研究交通的方法使用基于高性能计算架构的驾驶模拟器，实现高分辨率，基于物理的，逼真的交互式模拟，这对于在不断变化的交通条件下以及对自主，控制和连接的各种假设下对车辆和驾驶员进行详细建模以及监控人类驾驶行为是必要的。该项目识别、表征和塑造人类网络物理交互，这些交互必然会出现在未来的混合自动驾驶和传统车辆交通中。其目标是：(1)更好地理解混合交通中的人-网络-物理交互，特别是驾驶员对车辆自动化的信任；(2)开发分析和计算方法，以评估驾驶员从自动驾驶车辆模式切换到传统车辆模式（例如，在自动化场景中失去信任）的情景对交通容量和制度转换的影响；(3)开发工程解决方案，以增强网络与人类的互动和交通流量。本研究包括以人类行为分析和界面设计为基础的多尺度分析方法，混合交通和车辆控制建模，以及人、车辆和交通之间相互依赖的系统级分析。通过该项目了解人类网络与物理之间的相互作用，将有助于更好地设计自动驾驶汽车，以及人类与这些不断发展的系统安全可靠地互动的方式。该奖项反映了美国国家科学基金会的法定使命，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。

The promise of intelligent transportation systems is safety, mobility, and efficiencies well beyond current technologies. Yet, transportation engineers and planners must anticipate future transportation needs for which the enabling technologies are rapidly evolving while the change-over from old to new transportation technologies will take decades. This project seeks to understand the impact of these trends for transportation engineering, offer approaches for effective shared control between autonomous vehicles and drivers, and provide insight into strategies to realize the promise of intelligent transportation systems. The approach to studying traffic uses driving simulators, built on a high-performance computing architecture, that enables high-resolution, physics-based, realistic, interactive simulations, necessary for detailed modeling of vehicles and drivers under changing traffic conditions and various assumptions on autonomy, control, and connectedness, and for monitoring human driving behaviors.This project identifies, characterizes, and shapes human cyber-physical interactions, bound to emerge in future mixed autonomous and conventional vehicle traffic. Its aims are to: (1) better understand human-cyber-physical interactions in mixed traffic, particularly driver trust in vehicle automation; (2) develop analytical and computational methods to assess the impacts on traffic capacity and regime transition for scenarios in which a driver switches from autonomous vehicle mode to conventional vehicle mode (e.g., loss of trust in automation scenarios); and (3) develop engineering solutions to enhance cyber-human interactions and traffic flow. The research embraces a multi-scale analysis approach anchored in human behavior analysis and interface design, modeling of mixed traffic and vehicle control, and system-level analysis of interdependency among human, vehicles, and traffic. Understanding of human cyber-physical interactions through this project will enable better future designs of autonomous vehicles and the ways in which humans interact with these evolving systems safely and dependably.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.