In this presentation, we will explore an AI-enhanced approach to testing advanced driver assistance systems (ADAS) through simulations, focusing on open-loop and closed-loop methodologies. Open-loop simulations allow for validating system performance based on predefined inputs, while closed-loop simulations provide continuous feedback, enabling dynamic adjustment of system responses. We will highlight how AI is leveraged in both methodologies, specifically in end-to-end model training and evaluation, and for the generative creation of diverse testing scenarios. This integration of AI ensures a more robust and comprehensive testing framework, ultimately improving the reliability and safety of ADAS features in real-world applications.

• Key differences between open-loop and closed-loop simulations
• AI’s role in enhancing simulation
• AI's cost impact on ADAS development, testing and validation
• Generative scenario creation using AI