Conservative Estimation of Perception Relevance of Dynamic Objects for Safe Trajectories in Automotive Scenarios

• DOI: 10.48550/arxiv.2307.10873
• Publication Date: 2023-01-01
• Authors: Ken Mori; Kai Storms; Steven Peters
• Keywords: Relevance

Having efficient testing strategies is a core challenge that needs to be overcome for the release of automated driving. This necessitates clear requirements as well suitable methods testing. In this work, perception modules are considered with respect relevance. The concept relevance currently remains insufficiently defined and specified. paper, we propose novel methodology by exemplary application collision safety in highway domain. Using general system use case specification, corresponding derived. Irrelevant objects thus which do not limit set safe actions available ego vehicle under consideration all uncertainties. As an initial step, decomposed into functional scenarios For each scenario, possible both any other dynamic object formalized equations. constrained traffic rules, yielding criteria. result, present conservative estimation relevant need complete evaluation. provides applicable offline validation components. A visualization presented examples from highD dataset, showing plausibility results. Finally, possibility future outlined.