CARLA 0.9.2: Upgraded ROS-bridge, Traffic scenario engine and a new Agent class to perform driving navigation from client side!

We have continued extending CARLA's eco-system and today we are proud to bring one of the most expected features, Scenario_Runner. A new framework based on py_trees to generate traffic situations in CARLA, aiming to train and verify Autonomous Driving agents. We also bring un upgrade of the ROS-bridge compatible with 0.9.X, so that now you can run your ROS agents in CARLA. These new features among many other bug fixes and interesting additions

CARLA 0.9.1: Vehicle navigation, new waypoint-based API, maps creation and more

This new version brings back most of the missing functionalities of the 0.8.X branch. It also introduces a new programmable navigation API built on top of a waypoint-based representation based on ASAM OpenDRIVE. Users can now build their own maps by defining road networks in OpenDRIVE format. We have integrated VectorZero's RoadRunner to create new maps for CARLA in a semi-automatic way.