CARLA

CARLA is an open-source autonomous driving simulator built on Unreal Engine 4 for developing, training, and validating driving systems.

CARLA provides a high-fidelity environment for testing autonomous vehicle stacks across diverse urban layouts and weather conditions. Built on Unreal Engine 4, it utilizes the OpenDRIVE standard to define road geometries and offers a robust Python API for controlling traffic, pedestrians, and sensor suites (including LIDAR, cameras, and GPS). Researchers use CARLA to generate synthetic datasets and benchmark reinforcement learning agents in a risk-free, scalable sandbox. Its modular architecture supports integration with ROS and Autoware, making it the industry standard for bridging the gap between simulation and real-world deployment.

https://carla.org