Strategies based on Digital Twins and Software Defined Vehicles are echoing through every automotive industry hallway these days. Less prominently covered are the inherent – often missing – structural changes and tools required to support the development and validation steps of these strategies. In automotive radar, perception is crucial to the sensor’s value-add in ADAS/AD functionality. Underlying assumptions for the perception models used for functionality need to be validated against variants of physical sensor frontends, bumper designs, mounting positions, etc. Digital Twins of these variants are often missing, and field-testing is the measure of choice for quality and safety assurance of ADAS/AD functionality. We are now showing three major steps of perception model creation, robustness testing, and ground truth validation based on a pedestrian perception case study, using the unique Keysight, lab-based, over-the-air radar test environment. In this talk, you will:

• Discover the importance of structural changes and tools for supporting advanced automotive technologies
• Get a better understanding of the critical role of perception in AD systems
• Learn about the challenges in validating perception models against real-world factors