Table of Contents

SetUp

Minimum Hardware Requirements

We recommend the following PC operating environment for this tournament.

  • OS: Ubuntu 22.04
  • CPU: Intel Corei7 (8 cores) or higher
  • GPU: NVIDIA Geforce RTX 3080 (VRAM 12 GB) or higher
  • Memory: 32 GB or more
  • Storage: SSD 30 GB or higher

If you are unable to prepare a PC that meets the above specifications, please refer to the "For participants with two PCs" below.

For participants with two PCs

Autoware PC

  • OS: Ubuntu 22.04
  • CPU: Intel Corei7 (8 cores) or higher
  • GPU: NVIDIA Geforce GTX 1080 or higher
  • Memory: 16 GB or higher
  • Storage: SSD 10 GB or higher
  • For more information here

AWSIM PC

  • OS: Ubuntu 22.04 or Windows 10/11
  • CPU: Intel Corei7 (6 cores and 12 threads) or higher
  • GPU: NVIDIA Geforce RTX 2080 Ti or higher
  • For more information click here

Autoware and AWSIM PCs should be located in the same network. If they are located in the same network, topic communication between PCs is basically possible without any additional settings. In the unlikely event that topic communication is not possible, please disable the firewall or review the rules.

Environment Setup

AWSIM(Ubuntu)

Preparation

  • Install Nvidia drivers

    1. add repository
    sudo add-apt-repository ppa:graphics-drivers/ppa
    1. update package list
    sudo apt update
    1. install
    sudo ubuntu-drivers autoinstall
    1. After rebooting, run the following command to confirm that the installation has completed.
    nvidia-smi

    nvidia-smi

  • Install Vulkan

    1. update package list
    sudo apt update
    1. install libvulkan1
    sudo apt install libvulkan1

  • Prepare the course

    1. Download latest AWSIM_AIChallenge_Ubuntu_v*.*.zip from GoogleDrive and unzip it
    2. change permissions as shown in the figure
      Change the permissions as shown in the figure
      Double-click the file to launch it.
    3. confirm that the following screen is displayed awsim

Preparation of Docker

Docker image of Autoware (using CUDA) is available for this competition.

  • Preparation
    Please install the following.

  • Prepare and launch Docker image - Prepare Autoware

    1. get a Docker image
    docker pull ghcr.io/automotiveaichallenge/aichallenge2023-sim/autoware-universe-cuda:v1

    If the above method takes a long time or times out, you can use the following method.
    We have placed a tarball of the image at here. Please use the following command

    docker load < autoware-universe-cuda_v1.tar.gz
    1. download the data for the competition
    sudo apt install -y git-lfs
git lfs clone https://github.com/AutomotiveAIChallenge/aichallenge2023-sim
    1. build docker image for competition
    cd ./aichallenge2023-sim/docker
bash build.sh
    1. start rocker
    bash run_container.sh

Starting AWSIM in a Docker container

If you want to start AWSIM from a Docker container, please follow the steps below after installing a Docker image according to the Docker image preparation procedure (see below).

  1. extract the executable file for the convention in aichallenge2023-sim/docker/aichallenge (Hereinafter, it is assumed to be located in aichallenge2023-sim/docer/aichallenge/AWSIM/AWSIM.x86_64)
  2. launch the Docker container (please verify with docker container ls that container exists) Please confirm the existence of the following docker image using other terminal by using the docker image ls command. below is an example.
aichallenge-train        latest                            f5f05f758f55   2 weeks ago      14.9GB

Once confirmed, launch rocker with the following command.

 cd ./aichallenge2023-sim/docker
 bash run_container.sh

Open a new terminal and confirm the existence of the docker container by using the docker container ls command. below is an example.

CONTAINER ID   IMAGE          COMMAND       CREATED          STATUS          PORTS     NAMES
fdbe7cb05782   1f3d763bc501   "/bin/bash"   15 minutes ago   Up 15 minutes             elegant_hellman
  1. execute the following in the container
 sudo ip link set multicast on lo
 source /autoware/install/setup.bash
 /aichallenge/AWSIM/AWSIM.x86_64

AWSIM(Windows)

  1. Download latest AWSIM_AIChallenge_Windows_v*.*.zip from GoogleDrive and unzip it.
  2. double-click the file to start it Confirm that the following screen is displayed. awsim

Copy map data (pcd, osm)

mapfiles

Map data is stored in AWSIM compressed files. Copy the osm and pcd files located in AWSIM_Data/StreamingAssets/kashiwanoha2023_integ to aichallenge2023-sim/docker/aichallenge/mapfile and arrange them so that the file structure is as follows:

aichallenge2023-sim
└ docker
 └ aichallenge
  └ mapfile
   ├ .gitkeep
   ├ lanelet2_map.osm
   └ pointcloud_map.pcd

Autoware

  • Verify Autoware operation
    This section describes how to check the operation of Autoware using AWSIM.
    1. Start AWSIM.
    2. Start Autoware.
    # In the Rocker container
sudo ip link set multicast on lo
cd /aichallenge
bash build.sh
source aichallenge_ws/install/setup.bash
ros2 launch autoware_launch e2e_simulator.launch.xml vehicle_model:=golfcart sensor_model:=awsim_sensor_kit map_path:=/aichallenge/mapfile

    1. Confirm that the following screen (Rviz2) is displayed.
      autoware1

    2. Confirm that self-position estimation is done. If it is not estimated correctly, select 2D Pose Estimate in the tab and drag the actual position of the vehicle. autoware2

    3. Select the 2D Goal Pose in the Goal Pose tab and drag to specify the goal position. Confirm that the route is displayed and Routing changes from UNSET to SET as shown in the image (it takes a little time after you specify it). autoware3

    4. Press the AUTO button in OperationMode and confirm that self driving is started. autoware4

Edit this page