This year, at FKFS we completed our work on the publicly funded KI Data Tooling project. This project ran over a period of 45 months and was funded by the Federal Ministry for Economic Affairs and Climate Protection. In addition to FKFS, the consortium, led by BMW, included 16 other partners from industry and science.
In the KI Data Tooling project, a toolchain that can be used to generate realistic simulations with multimodal sensor sets from camera, lidar and radar, was developed. In particular, FKFS developed a multi-simulation handler for the configuration of co-simulations, which can be used to link a wide range of data generation tools. We rely on standardized formats such as OpenSCENARIO, OpenDRIVE, OSI, etc. Furthermore, our project team at FKFS looked at the influence of driving dynamics on sensor effects and the generation of realistic traffic scenarios. Another of our focuses was on the implementation of a real-time capable toolchain for use in the Stuttgart driving simulator, in which driver-in-the-loop simulations can be carried out.
For us at FKFS, the results of the KI Data Tooling project represent significant progress in realistic driving simulation and the application of neural networks with a focus on automated driving.