NSF-BSF: HCC: Cultural Differences in Pedestrian-Autonomous Vehicle Interaction

行人通过他们的动作、言语交流、肢体语言和目光与司机进行含蓄和明确的交流和协商。道路使用者之间的认知、理解和行动的不匹配很容易引发事故。自动驾驶的出现将需要对车辆和行人互动的复杂方式有更明确的理解。这项研究旨在模拟驾驶员如何在道路上与行人进行隐式沟通和协调，并评估这些驾驶互动在不同文化中的差异。这项研究的科学贡献将是驾驶交互的新模型，该模型描述了道路使用者之间发生的隐式和显式沟通和协商模式。 这项研究对于使自动驾驶车辆能够感知、响应和适应区域道路交互规范，以及提高自动化的安全性和信任度至关重要。这项工作通过在开发先进技术时考虑不同的文化，以减少与不良设计、文化偏见或不必要的道路测试相关的风险，从而造福社会。这项研究的科学贡献将是驾驶交互的新模型，该模型描述了道路使用者之间发生的隐式和显式沟通和协商模式。该项目包括对美国、以色列和丹麦行人与车辆交互的自然数据的收集和分析，使用虚拟现实模拟进行实验研究，使用受控场景来引发行人和驾驶员之间的交互行为，并根据收集的数据开发行人与驾驶员的情境感知模型。该团队将积极吸引汽车行业和学术研究人员来验证模型并创建行人与车辆交互行为的全球数据库。该奖项反映了 NSF 的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

Pedestrians communicate and negotiate with drivers implicitly and explicitly through their movement, as well as verbal communication, body language, and gaze. Mismatches in perception, understanding and action between road users can easily cause accidents. The advent of autonomy will necessitate a more explicit understanding of the complex manner in which vehicles and pedestrians interact. This research seeks to model how drivers implicitly communicate and coordinate with pedestrians on the road, and to assess how these driving interactions differ across cultures. The scientific contribution of this research will be new models of driving interaction that characterize the implicit and explicit communication and negotiation patterns that occur between road users. This research is critical to enabling autonomous vehicles to sense, respond and adapt to regional road interaction norms, and to improve safety and trust in automation. This work benefits society by considering different cultures in the development of advanced technologies in order to reduce risks associated with poor design, cultural bias or unnecessary on-road testing.The scientific contribution of this research will be new models of driving interaction that characterize the implicit and explicit communication and negotiation patterns that occur between road users. The project incorporates the collection and analysis of naturalistic data of pedestrian-vehicle interactions in the US, Israel and Denmark, the use of virtual reality simulation for experimental study using controlled scenarios with which to elicit interactive behaviors between pedestrians and drivers and the development a pedestrian-driver situation awareness model from the data collected. The team will actively engage automotive industry and academic researchers to validate models and to create a global database of pedestrian-vehicle interaction behavior.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.