A Modular Framework for Virtual Calibration and Validation of Driver Assistance Systems

Along with the electrification and connectivity of vehicles, the automation of the driving task represents one of the main trends in the automotive industry. With the increasing capabilities of Advanced Driver Assistance Systems (ADAS) and Automated Driving Systems (ADS), the number of driving scenarios that must be handled by the systems in road traffic is constantly increasing. Drivers expect the vehicle to act safely, comfortably and efficiently, which is why ADAS/ADS must successfully perform a calibration and validation process during vehicle development. While this process is nowadays accomplished on proving grounds or in road traffic, a variety of new approaches will be required in the future to deal with the increasing scope of the systems. One of these approaches is the use of simulation tools to perform virtual calibration and validation. The main challenge of this approach is the multitude of components, such as a scenario catalog, the simulation itself, evaluation metrics and more, that are required in high quality to generate usable results. To meet this challenge, the paper presents a modular framework for both virtual calibration and validation of ADAS/ADS. The focus is on the presentation of the modules required for a combined approach as well as their structure and interfaces. The paper first presents the framework in a generalized structure that can be applied to a wide variety of ADAS/ADS, and then shows the exemplary implementation for virtual calibration and validation of an adaptive cruise control. Finally, the results of a virtual calibration and validation run are shown and discussed.