灵活性(工程)
计算机科学
仲裁
吞吐量
功能(生物学)
频道(广播)
运输工程
风险分析(工程)
工程类
无线
计算机网络
电信
业务
统计
数学
进化生物学
政治学
法学
生物
作者
C.A.J. Hanselaar,Emilia Silvaş,Andrei Terechko,W.P.M.H. Heemels
出处
期刊:
日期:2022-10-08
被引量:2
标识
DOI:10.1109/itsc55140.2022.9922405
摘要
Autonomous vehicles (AVs) promise to reduce greenhouse gas emissions, increase comfort and throughput of transportation, while simultaneously significantly reducing traffic deaths. To allow autonomous vehicles to satisfy the high safety levels for unsupervised participation in realistic traffic, the faults and functional insufficiencies of AV systems need to be mitigated during operation. Unfortunately, current state-of-the-art functional insufficiency detection and mitigation methods do not provide large enough safety improvements, without impeding the availability of safe autonomous functionality. To fill this gap, we propose the Safety Shell, an implementable multi-channel architecture and arbitration method. The ability to increase the number of parallel AV function channels allows for a path to safe AV systems, while the novel arbitration method ensures availability for comfortable journey continuation. The flexibility and benefits of the Safety Shell are shown using use case studies.
科研通智能强力驱动
Strongly Powered by AbleSci AI