A Needle in a Haystack -- How to Derive Relevant Scenarios for Testing Automated Driving Systems in Urban Areas

• DOI: 10.48550/arxiv.2109.03648
• Publication Date: 2021-01-01
• Authors: Nico Weber; Christoph Thiem; Ulrich Konigorski
• Keywords: Toolchain; Haystack; Scope (computer science); Scenario testing

While there was great progress regarding the technology and its implementation for vehicles equipped with automated driving systems (ADS), problem of how to proof their safety as a necessary precondition prior market launch remains unsolved. One promising solution are scenario-based test approaches; however, is no commonly accepted way systematically generate extract set relevant scenarios be tested sufficiently capture real-world traffic dynamics, especially urban operational design domains. Within scope this paper, overall concept novel simulation-based toolchain development testing ADS-equipped in environments presented. Based on previous work highway environments, developed enhancements aim at empowering able deal increased complexity due more complex road networks multi-modal interactions various participants. derived requirements, thorough explanation different modules constituting given, showing first results identified research gaps, respectively. A closer look taken two use cases: First, it investigated whether capable serve synthetic data source within phase enrich scenario database terms extent, impacts what-if-scenarios future mixed traffic. Second, analyzed combine individual advantages real recorded an agent-based simulation so-called adaptive replay-to-sim approach support vehicle. The contributes overarching goal methodology validation vehicles, environments.