首页> 外文期刊>Journal of Risk and Financial Management >Customer Preferences and Implicit Tradeoffs in Accident Scenarios for Self-Driving Vehicle Algorithms
【24h】

Customer Preferences and Implicit Tradeoffs in Accident Scenarios for Self-Driving Vehicle Algorithms

机译:自驾车算法在事故场景中的客户偏好和隐式权衡

获取原文
           

摘要

The development of self-driving vehicles is proceeding rapidly and with significant investment of resources. However, a full-scale deployment is not imminent. Among the challenges self-driving vehicles are facing, they will have to navigate complex ethical challenges. The algorithms governing their behavior will have to decide how to steer them in situations where accidents cannot be avoided. In some of these situations they will have to decide which of several potential parties to injure in the process. We investigate the preferences of Swiss customers for this decision by forcing a selection between simplified scenarios where a given number of car passengers or a given number of pedestrians will be killed in the accident. Both passengers and pedestrians can be adults or children. The passengers are explicitly identified as the respondent themselves and their family. While children are implicitly valued higher than adults, Swiss customers value passengers and pedestrians implicitly roughly equally, and assign increasingly higher marginal values to additional people, both passengers and pedestrians. These results seem to partially contradict similar studies conducted in other countries and recent statements by automotive companies, potentially indicating the need to adapt both corporate communications and steering algorithms in different geographies.
机译:无人驾驶汽车的发展迅速,并且投入了大量资源。但是,全面部署并非迫在眉睫。在自动驾驶汽车面临的挑战中,他们将不得不应对复杂的道德挑战。控制其行为的算法将必须决定在无法避免事故的情况下如何引导它们。在某些情况下,他们将不得不确定在此过程中可能伤害的几个潜在方中的哪一方。我们通过在简化方案中进行选择来调查瑞士客户对这一决定的偏好,在简化方案中,事故中将杀死一定数量的汽车乘客或一定数量的行人。乘客和行人都可以是成年人或儿童。乘客被明确确定为被访者本人及其家人。尽管儿童的隐性价值要比成人高,但瑞士顾客对乘客和行人的价值却大致相同,并为其他人(乘客和行人)赋予越来越高的边际价值。这些结果似乎与在其他国家/地区进行的类似研究以及汽车公司的最新声明相矛盾,这可能表明有必要在不同地区同时采用公司通信和转向算法。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号