Strategies for Negotiation between Autonomous Vehicles and Pedestrians
Strategies for Negotiation between Autonomous Vehicles and Pedestrians
复制标题
自动驾驶车辆和行人之间的谈判策略
DOI：
--
复制
发表时间：
2015
期刊：
影响因子：
--
通讯作者：
A. Riener
中科院分区：
文献类型：
--
作者：
F. Keferböck;A. Riener
自动驾驶汽车似乎即将成为现实，汽车制造商展示了他们的第一个原型车，大众媒体也开始讨论自动驾驶汽车的话题。到目前为止，在个人交通中，人们使用私家车或摩托车、骑自行车或步行（从最强到最弱的方式）从 A 到 B 行驶距离。所有这些模式在典型的交通情况下并行共存，并且应该清楚的是，不同的情况需要不同道路参与者之间的澄清和沟通，例如，协商谁拥有通行权以及谁必须等待。许多司机每天都会花相当多的时间在车里——上下班、购物和旅行。为了节省驾驶员的时间，预计手动驾驶将在不久的将来被淘汰并被自动化系统取代。迄今为止，自动驾驶汽车制造商尚未提出的问题之一是，当“最强”的道路使用者（车辆或卡车）不再由人类驾驶时，脆弱的道路使用者（VRU）沟通、互动和谈判的机会也可能被驱逐。在这项工作中，基于今年 CES 上梅赛德斯·奔驰 F015 的展示，我们希望表明，用自动驾驶汽车替代行人与车辆之间的通信方式来理解行人的手势和手势，并与他们积极沟通（例如，使用挡风玻璃、发动机盖或前灯上的视觉反馈）非常重要。为了更深入地了解这一场景，我们将设置并进行用户研究，将受试者置于可能是自动驾驶汽车的情况下，并比较汽车与受试者明确交互和不与受试者明确交互的情况下的动作和反应。我们的期望是检测行人行为的差异，这将揭示出对自动驾驶汽车不同程度的信任和信心。
Autonomous cars seem on the verge to reality, with vehicle manufacturers presenting their first prototypes and the topic of self-driving vehicles being discussed in mass media. Until now, in individual traffic humans covered distances from A to B using a personal car or a motorcycle, riding a bike or walking by foot (from strongest to weakest modality). All these modalities coexist in parallel in typical traffic situations, and it should be clear that different situations require clarification and communication between the different road participants, e.g., to negotiate who has right of way and who has to wait. Many drivers spend a considerable time each day in their car – for commuting, shopping, and traveling. In order to save time for the driver it is expected that manual driving will be eliminated in the near future and replaced by automated systems. One of the problems not brought up by autonomous vehicle manufacturers so far is when the „strongest“ road user (vehicle or truck) is no longer human-driven, as then the chance for vulnerable road users (VRUs) to communicate, interact and negotiate could be evicted too. In this work, based on the showcase of Mercedes Benz’s F015 at CES this year, we want to show that it is important to substitute the means of pedestrian-vehicle communication by autonomous cars to understand the signs and gestures of pedestrians and also communicating actively (e.g., using visual feedback on windscreen, bonnet or headlights) towards them. To get a deeper knowledge of this scenario, we will setup and conduct a user study, placing subjects into situations with a presumably autonomous car and comparing the actions and reactions with and without the car explicitly interacting with the subject. Our expectation is to detect a difference in the behavior of the pedestrians that will reveal a different level of trust and confidence towards autonomous cars.