This presentation highlights recent research on safety algorithms that enhance autonomous vehicle decision-making in real-time. Focusing on obstacle avoidance, pedestrian interactions, and takeover situations, we explore how advanced logic and AI models work together to ensure safe responses in complex environments. Results from both simulation and real-world tests reveal key insights for improving AV safety and readiness for deployment.

In this session, you will learn more about:

• Improved obstacle and pedestrian handling using hybrid safety algorithms
• Critical considerations for safe and timely vehicle takeovers
• Scenario-based testing as a foundation for validating AV safety functions