Advanced driver assistance systems (ADAS) are becoming increasingly important for today's commercial vehicles. It is therefore crucial that different ADAS functionalities interact seamlessly with existing electronic control unit (ECU) networks. For example, autonomous emergency braking (AEB) systems directly influence the brake ECU and engine control. It has already become impossible to reliably validate this growing interconnectedness of control interventions in vehicle behavior with prototype vehicles alone. The relevant tests must be brought into the lab at an earlier development stage to evaluate ECU interaction automatically.
This paper presents an approach for using hardware-in-the-loop (HIL) simulation to validate ECU networks for extremely diverse ADAS scenarios, while taking into account real sensor data. In a laboratory environment, the sensor systems based on radars, cameras, and maps are stimulated realistically with a combination of simulation and animation. This makes it possible to validate ADAS functionalities, such as lane departure warning (LDW) systems, AEB and predictive cruise control (PCC), already in the lab.
The project described in this paper is a cooperation of MAN Truck & Bus AG and dSPACE GmbH. To simulate realistic driving environment scenarios, several dSPACE Automotive Simulation Model (ASM) components with special enhancements are combined to create a virtual truck. By using various real sensors simultaneously, new possibilities for investigating sensor data fusion are offered even before a real prototype exists. Specially designed automatic monitoring of the real cockpit instrumentation provides additional validation options, further increasing test maturity. In the final development phase, the laboratory results for driver assistance systems are validated by performing actual test drives.
