Martin Herrmann is Senior Business Development Manager at IPG Automotive in Frankfurt am Main. After studying automotive engineering at Ilmenau Technical University, he joined IPG Automotive in Karlsruhe in 2014, initially leading ADAS and vehicle-dynamics engineering projects.
Since 2016, Martin has focused on consulting and designing solutions for simulation-based development and testing of automated-driving functions, specialising in scenario-based testing and data-driven development.
Throughout his career, he has contributed to major industry initiatives, including the PEGASUS and SET Level research projects.
Presentation
From Real-World Replay to Closed-Loop Validation
Recorded sensor data is a valuable foundation for ADAS and automated driving validation. Replay enables reproducible testing with real-world data and is widely used to evaluate perception performance, regression behavior, and system robustness. However, replay alone cannot assess how a system reacts to changing situations or validate its behavior in a closed loop. This presentation shows how sensor data injection extends recorded test drives towards interactive validation, enabling additional test cases focused on decision-making, vehicle behavior, and system reactions. A common hardware platform supporting both open-loop replay with recorded sensor data and closed-loop testing with synthetic sensor inputs is presented. Replay and closed-loop validation are highlighted as complementary methods with different strengths and validation objectives that together provide broader and more efficient test coverage.
