Creating detailed road maps is essential for autonomous driving systems, as they serve as a safeguard by adding redundancy to the vehicle’s understanding of the environment when sensor data becomes limited or compromised. However, generating these maps can be resource-intensive due to the high level of complexity and precision required, and the use of varied formats across organizations further hinders interoperability.

To address the critical need for interoperable road maps in ADS, we present the first publicly available converter OpenCat, which converts the OpenDRIVE format to the Catmull-Rom spline representation. OpenDRIVE is a widely used format for high-fidelity simulation as it supports detailed map data, while the Catmull-Rom Spline enables realistic and continuous paths that are useful for trajectory planning. We assess OpenCat’s performance by applying it to the SensoDat dataset, which is publicly available, and exhibit that our method effectively converts OpenDRIVE geometry into testable and realistic road scenarios. Furthermore, we successfully used the converted roads for the Udacity self-driving car simulator, demonstrating the practical applicability of the converted roads in simulation environments.