The Scenario-in-the-Loop (SciL) automotive simulation concept and its realisation principles for traffic control

Abstract

As autonomous and driver assistance functions become more and more complex, novel testing methodologies are required. Introduction of highly automated and autonomous vehicles on public roads alters traffic flow dynamics. Therefore, it is not enough to test only the single vehicle itself, but it will also be required to test its interaction with the environment and other vehicles of traffic. In this context, a practical goal is to test independently of real traffic as much as possible. The Vehicle-in-the-Loop method (i.e. where the vehicle is the ’hardware’) is already used for testing vehicles. In this configuration the vehicle is directly forced to do specific manoeuvres, but the sensors of the vehicle are not used and the vehicle does not have to make decisions. The concept of the Scenario-in-the-Loop introduces a testing method where the physical attributes as well as the sensors of vehicles are tested via a virtual twin. In the paper we present the main principles of the Scenario-in-the-Loop concept and outline the requirements of such a test environment from the perspective of traffic control.