备份
计算机科学
过程(计算)
光学(聚焦)
计算机辅助设计
人机交互
风险分析(工程)
工程类
数据库
业务
工程制图
操作系统
光学
物理
作者
Frederik Naujoks,Dennis Befelein,Katharina Wiedemann,Alexandra Neukum
出处
期刊:Advances in intelligent systems and computing
日期:2017-06-24
卷期号:: 525-537
被引量:53
标识
DOI:10.1007/978-3-319-60441-1_52
摘要
Conditionally automated driving (CAD) relieves the driver from monitoring current traffic conditions. This type of driving inherently enables the driver to execute different non-driving-related tasks (NDRTs). However, the driver still must be available as a backup option. With this in mind, the classification and evaluation of various NDRTs concerning their impact on driver performance in takeover scenarios represent an important contribution toward the creation of safe CAD functions. We reviewed various NDRTs that were used in studies on automated driving. The focus was on assigning aspects of these activities (e.g., ability to visually monitor traffic, necessity of sustained attention to NDRT, etc.) to various steps of the takeover process (e.g., noticing and interpreting takeover requests), which could be impaired by the execution of the respective NDRT. This, in turn, would increase the demands on the driver with respect to managing the takeover situation.
科研通智能强力驱动
Strongly Powered by AbleSci AI