Use of motion capture to secure automated driving functions

Efficient methods for the development and validation of highly automated vehicle systems are playing an increasingly important role in the mobility industry. In particular, the introduction of innovative assistance and automation functions requires new methods in simulation and test bench technology. In this paper, a new approach for the development of a virtual test environment for highly automated vehicles based on a virtual reality engine is presented. The basis is a realistic, georeferenced city model, which is reactively connected in co-simulation with a highly detailed vehicle model on a vehicle-in-the-loop test bench. Models of sensor systems such as camera and radar allow the vehicle to interact in real time with the environment and road users. These VRUs (Vulnerable Road Users) are based on motion capture technology and are integrated as realistic avatars. The efficient use of this technology for high-resolution motion capture of VRUs such as pedestrians, cyclists and motorcyclists and the integration as digital twins into AVL's simulation and test bench environments is presented. © 2023 The Authors.