The SET Level project, carried out by the Institute for Automotive Engineering (ika) at the Rheinisch-Westfälische Technische Hochschule Aachen, was a follow‑up to the PEGASUS programme (2016‑2019) and focused on the use of simulation for the development and testing of highly automated driving functions in urban environments. The project ran until 31 October 2022 and was funded by the German Federal Ministry for Economic Affairs and Climate Action (BMWK) under grant number 19A19004C. It brought together automotive manufacturers, suppliers, IT vendors and research institutions, with the ika contributing the bulk of the effort in subproject 3, while subproject 2 received a smaller share of resources.

In subproject 2 the team concentrated on system architecture, scenario definition and model integration. A key outcome was the extension of the OpenSCENARIO standard and the development of an interface architecture that allows traffic agents to be controlled via the OpenSCENARIO and OSI standards. The resulting architecture supports scalable simulation of traffic scenarios and enables the automatic generation of road layouts for required variations. The work also produced a traffic‑agent model architecture (UAP 2.1.1) and interface extensions (UAP 2.1.4 and 2.1.5) that integrate traffic agents into the overall simulation environment.

Subproject 3, where the ika’s main contribution lay, delivered a closed‑loop simulation framework for automated driving functions. The team developed and calibrated driver and road models, extending them to support automatic lane and intersection generation. The models were specified, packaged as Functional Mock‑up Units (FMUs) and integrated into the CARLA simulation platform. A state‑diagram representation of the base manoeuvre model and a schematic of the GPS‑model concept were produced, demonstrating how input and output data are handled according to the OSI standard. Test results from the closed‑loop agent model show that the simulation can reproduce realistic vehicle dynamics and interaction with traffic participants, providing a robust basis for validation and qualification of automated driving functions.

The project also addressed data management for simulation models, defining specifications for parameter‑set and metadata handling. The ika’s work on model selection and integration introduced mechanisms to identify suitable models for specific use cases and to embed them seamlessly into the simulation workflow. Throughout the project, the team performed validation and qualification tasks to ensure that the models behaved correctly both individually and in combination.

In addition to the technical deliverables, the project produced several publications and contributed to the development of a scenario description language that aligns closely with the ASAM OpenSCENARIO standard. The collaboration with industry partners facilitated the exchange of best practices and the adoption of the developed tools and models across the automotive sector. The overall effort, amounting to 7 person‑months in subproject 2 and 47 person‑months in subproject 3, demonstrates the feasibility of a simulation‑centric approach to the development of highly automated driving functions in complex urban settings.