Realistic vehicle trajectories and driving parameters from CARLA autonomous driving simulator

This dataset contains realistic trajectories from multiple vehicles moving in the simulated environment of CARLA autonomous driving simulator. Two different maps (Map04 and Map10) have been exploited, corresponding to realistic driving conditions in simulated urban environments. Five sub-datasets have been extracted, corresponding to different number of vehicles, e.g., 50, 100 and 200, spawned in each map for 200 seconds. Every record of the dataset consists of: time stamp, vehicle ID, ground truth position x, y, z in CARLA reference system, ground truth pitch, roll and yaw angles (in degrees), ground truth linear velocity in x, y and z direction (m/s), ground truth linear acceleration in x, y and z direction (m^2/s) and ground truth angular velocity in x, y and z direction (deg/s).