The validation of autonomous driving systems is challenged by the growth of scenarios resulting from complex interactions between vehicles, traffic participants, infrastructure and environmental conditions. Exhaustive real-world testing or brute-force simulation is not feasible. This presentation introduces a combined physics-based, traffic-aware and AI-driven simulation approach to manage scenario explosion. High-fidelity traffic modeling enables realistic multi-agent interactions and emergent behaviors, while AI techniques guide scenario generation and prioritization based on safety and performance KPIs. Integrated into closed-loop SIL and HIL workflows, this approach increases coverage of critical traffic situations while reducing validation effort, supporting safety and regulatory requirements.

• Traffic interactions drive scenario complexity in ADAS and AV validation
• High-fidelity traffic modeling captures multi-agent and emergent behaviors
• AI prioritizes traffic-critical scenarios based on safety and performance KPIs
• Physics-based closed-loop SIL/HIL simulation ensures credibility and traceability
• Efficient scenario coverage supports safety standards and future homologation