As follows:
$ xhost local:
$ docker run --env DISPLAY=${DISPLAY} -v <<DATA_DIR>>:/home/autoware/data -v /tmp/.X11-unix:/tmp/.X11-unix:ro --rm onohr/autoware
autoware@autoware:~ <<SET UP DATASET in DATA_DIR>>
...
autoware@autoware:~$ Autoware/ros/run
...
autoware@autoware:~$ exit
You need to install Docker Plugin provided by NVIDIA in order to access NVIDIA GPUs from Docker Container.
Please see the Quickstart (https://github.com/NVIDIA/nvidia-docker/#quickstart).
Execute the following command.
$ make AUTOWARE_VERSION=1.7.0 dl_autoware
$ make AUTOWARE_VERSION=1.7.0 SERVICE_NAME=cudagl clean build
$ make AUTOWARE_VERSION=1.7.0 SERVICE_NAME=opengl clean build
default is ...
$ make AUTOWARE_VERSION=1.7.0 SERVICE_NAME=opengl sh
or
$ make AUTOWARE_VERSION=1.7.0 SERVICE_NAME=cudagl sh