Managing complex driving situations is a non-trivial task for autonomous traffic agents. To guarantee desirable system properties of an agent’s driving decisions, formal methods offer a wide range of approaches. In previous work, we introduced Urban Multi-lane Spatial Logic (UMLSL) to prove the safety of local driving decisions of Autonomous Traffic Agents (ATAs), e.g. for turning left at an intersection. We now propose to integrate UMLSL into a Traffic Game model, allowing us to verify properties of driving strategies. We use Strategy Logic to compare different driving strategies w.r.t. diverse metrics, e.g. energy-consumption or timeliness. We currently intensify our work on extensions and applications of the UMLSL universe and we also reflect on some recent and future challenges for that in this paper.

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

It’s Safe to Play While Driving: From a Spatial Traffic Logic Towards Traffic Games

  • Maike Schwammberger,
  • Qais Hamarneh

摘要

Managing complex driving situations is a non-trivial task for autonomous traffic agents. To guarantee desirable system properties of an agent’s driving decisions, formal methods offer a wide range of approaches. In previous work, we introduced Urban Multi-lane Spatial Logic (UMLSL) to prove the safety of local driving decisions of Autonomous Traffic Agents (ATAs), e.g. for turning left at an intersection. We now propose to integrate UMLSL into a Traffic Game model, allowing us to verify properties of driving strategies. We use Strategy Logic to compare different driving strategies w.r.t. diverse metrics, e.g. energy-consumption or timeliness. We currently intensify our work on extensions and applications of the UMLSL universe and we also reflect on some recent and future challenges for that in this paper.