Carla Starter Project for version 0.9.13

This project contains a multi-container Carla project template for educational purposes.

Setup the environment

• Ubuntu 20.04 environment (Ubuntu on WSL may work, not recommended)
• Docker (first install Docker and then follow Docker post-installation steps)
• Docker Compose (install docker-compose)
• Python 3.8 (comes by default in Ubuntu 20.04)
• Install and upgrade pip to the latest version
• Install carla package using pip install carla==0.9.13

Quickstart

Once your environment set up, start Carla and Carlaviz server from the project directory using

docker-compose --profile carlaviz up -d 

Then visit the address localhost:8080 in your browser to see a web-based visualization of the Carla simulation, which uses the default map. You can find an example client script in the project directory and run it using

python client/run.py

and close servers once done using

docker-compose down

Changing the map before the simulation

The carlaviz container does not support change of maps at runtime (See the related issue). Follow the steps to change the map in your simulation.

docker-compose up -d
python utils/config.py --map <CARLA_MAPNAME>
docker-compose --profile carlaviz up -d 
python client/run.py

where <CARLA_MAPNAME> denotes a Carla map name Town01, ..., Town07, Town10.

Don't forget to visit localhost:8080 to see the visualization.