Ensuring the safety of AVs remains a critical challenge, requiring both rigorous engineering methodologies and innovative testing approaches. This presentation explores how human-centered design principles can be leveraged to engineer safe autonomous driving (AD) systems and how large language models (LLMs) can enhance safety validation through automated test generation. By combining human expertise with AI-driven testing, we can improve scenario coverage, accelerate safety assessments, and align AV development with regulatory expectations. Attendees will gain insights into integrating LLMs into safety workflows and learn how to optimize the collaboration between human engineers and AI-driven tools to build trustworthy AV systems.
In this session, you will learn more about:
- Leveraging human expertise to define safety-critical scenarios and validate AV decision-making
- Using LLMs to generate diverse and challenging test scenarios, enhancing coverage and efficiency in AV validation
- Aligning AI-enhanced safety validation with industry standards such as ISO 26262, UL 4600, and ISO/TS 5083 for regulatory acceptance
&nsbp;
