当前位置: X-MOL 学术Cognit. Tech. Work › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Functional safety analysis of SAE conditional driving automation in time-critical situations and proposals for its feasibility
Cognition, Technology & Work ( IF 2.6 ) Pub Date : 2020-10-10 , DOI: 10.1007/s10111-020-00652-x
Takayuki Hirose , Tetsuo Sawaragi , Hideki Nomoto , Yasutaka Michiura

The autonomous driving technology is rapidly developed for commercial use, aiming at the conditional driving automation or the third level of driving automation (LoDA 3). One of the most critical challenges for the achievement is the smooth authority transfer from the system to human drivers in an emergency. However, it is still on the way to find out fundamental solutions. The difficulty is closely related to the envisioned world problem, for which the concept of functional modeling could be a solution. This paper presents a safety analysis of the authority transfer problem based on the ideas of functional modeling. We discuss the safety of the authority transfer in a time-critical situation by using a simulator based on the functional resonance analysis method (FRAM). The result shows that the involvement of human drivers in driving activities is still essential even during autonomous driving. We show that the current LoDA 3 is a myth that the human drivers are supposed to be required just in non-normal situations, while they can be free from dynamic driving tasks (DDTs) in usual cases. Based on the result, this work makes some proposals for successful autonomous driving, effective human–machine collaboration, and the right design of artifacts.

中文翻译:

时间紧迫情况下 SAE 条件驾驶自动化的功能安全分析及其可行性建议

自动驾驶技术快速商用,瞄准条件驾驶自动化或第三级驾驶自动化(LoDA 3)。实现这一目标的最关键挑战之一是在紧急情况下将权力从系统顺利转移到人类驾驶员。然而,它仍在寻找根本解决方案的路上。困难与预想的世界问题密切相关,功能建模的概念可以作为解决方案。本文基于功能建模的思想对权限转移问题进行了安全分析。我们通过使用基于功能共振分析方法 (FRAM) 的模拟器来讨论在时间紧迫的情况下权限转移的安全性。结果表明,即使在自动驾驶过程中,人类驾驶员参与驾驶活动仍然必不可少。我们表明,当前的 LoDA 3 是一个神话,认为人类驾驶员应该只在非正常情况下才需要,而在通常情况下,他们可以摆脱动态驾驶任务 (DDT)。基于结果,这项工作为成功的自动驾驶、有效的人机协作和正确的工件设计提出了一些建议。
更新日期:2020-10-10
down
wechat
bug