CARLA Simulator

CARLA (Car Learning to Act) is an open-source simulator designed to support research and development in autonomous driving systems. It provides a highly realistic and flexible environment for testing algorithms in perception, planning, and control without the constraints and risks of real-world deployment. Built on Unreal Engine, CARLA offers photorealistic rendering, dynamic weather conditions, and complex traffic scenarios, enabling researchers to validate solutions under diverse conditions.

By integrating CARLA into CCAM initiatives, we can accelerate innovation in connected and automated mobility. CARLA’s modular architecture allows interoperability with external tools, including ROS and OpenDRIVE, making it ideal for collaborative projects. Its open-source nature ensures transparency, reproducibility, and cost-effectiveness, aligning perfectly with Horizon Europe’s principles of open science and shared resources.

CARLA is already being successfully used in major Horizon Europe projects such as SUNRISE, BERTHA, and AI4CCAM, demonstrating its maturity and relevance for large-scale research and innovation. These projects leverage CARLA to simulate cooperative driving strategies, V2X communication protocols, and safety-critical scenarios, proving its value as a common platform for experimentation.

We propose offering CARLA as a shared simulation environment for CCAM partners to foster synergies between industry, academia, and public authorities. This approach will reduce development time, enhance trust in automated systems, and strengthen Europe’s leadership in sustainable and safe mobility.