The Role of Digital Twins in Automotive R&D for Rapid Prototyping and System Integration

Abstract

The automotive industry is undergoing a significant transformation as the focus shifts toward connecting, autonomous, shared, and electric vehicles (CASE). These technologies promise substantial improvements in vehicle safety, efficiency, and comfort, but achieving higher levels of automation poses new challenges for deep learning-based algorithms in driver assistance and automated driving functions. Validation and verification remain challenging due to the difficulty of generating and executing a sufficient quality of ground truth for all possible corner cases. The challenges are exacerbated when the vehicles are expected to learn and adapt to highly dynamic environments or when the validation is to be done in meaningful real-world scenarios. Simulation environments are required to generate ground truth, and their reliability must be assessed before evaluating the functionality of the AD systems within that simulation. The approach presented is called “Digital Twins of Vehicles and Driving Environments.”