DistanceAssistant is a social distancing monitor which visually alerts people when they are not observing proper social distancing guidelines. The application is packaged and deployed as a docker container.
Inside of the Docker container, this application uses the Robot Operating System (ROS) as the underlying framework for runtime, launch, and configuration. For more details, see: https://www.ros.org/
This project has been run/tested on a machine with the following specifications/software installed:
- Ubuntu 18.04
- Other operating systems may also work, but exposing the RealSense camera USB to VMs is incredibly tricky. Thus, a Linux OS with native Docker support is recomended.
- An Nvidia GPU
- The GPU must support CUDA as mentioned below
- We recommend a GTX 1070 or better
- CUDA 10.0 or greater: https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html
- An Intel DS435i Realsense camera
- Other models may work, but this is the only model that was tested
- docker-ce
- Tested on Docker 19.03, but older/newer versions may work
- nvidia container toolkit
- Intel RealSense SDK
- Intel realsense drivers can be installed by following the directions here: https://github.com/IntelRealSense/librealsense/blob/master/doc/installation.md.
- Note: the pre-built dkms drivers provided by Intel did not work for us. We compile the drivers using the instructions in the librealsense README for compilation.
NOTE: It is recommended that the RealSense camera be attached to the PC via a USB 3.1 port with a powered USB hub or an "active" (powered) USB cable. During testing on various PCs, we have seen intermittent issues due to lack of power on the USB port. This can happen across two PCs with the same model number. A powered USB3 hub greatly improves the reliability of the RealSense camera
To follow along with an example host setup from a fresh installation of Ubuntu 18.04, see the Example Host Setup Instructions.
The following instructions use the --gpus=all
flag to enable the NVIDIA GPU inside of the Docker container. However, for older versions of Docker (pre-19.03), the flag would be --runtime=nvidia
. See the NVIDIA Container Toolkit instructions above for more.
Ubuntu 18.04 Alternative Method
An alternative method to passing the --gpus
flag would be to modify the /etc/docker/daemon.json
file to use NVIDIA as the default runtime.
To do so, install the nvidia-container-runtime:
sudo apt install nvidia-container-runtime
Then, add the following to your /etc/docker/daemon.json
:
"default-runtime": "nvidia",
"runtimes": {
"nvidia": {
"path": "/usr/bin/nvidia-container-runtime",
"runtimeArgs": []
}
}
Finally, restart Docker:
sudo service docker restart
Now, you can omit the --gpus
flag from the runtime Docker command.
To build the docker image, first download the yolo model weights and build the docker file. The weights should be stored in the distance_assistant/src
folder.
# cd <distance-assistant folder>
# The below command assumes that it's run in the root folder of the
# repository
wget https://github.com/AlexeyAB/darknet/releases/download/darknet_yolo_v3_optimal/yolov4.weights -P distance_assistant/src/
docker build --network=host . -t distance_assistant/prototype
# The source tree should look like this:
# <root>/
# distance_assistant/
# scripts/
# msg/
# src/
# yolov4.weights
# Dockerfile
# README.md
We suggest to mount the camera between 1.7 and 2.1 meters above the ground and slightly tilted down (~10 degrees) in order to maximize the usable depth sensor FOV.
During the application startup we perform an automated calibration step which computes the camera pitch & roll as well as it's height from the ground plane. Therefore it's important that the camera is stationary and there are no large objects in front of the camera blocking the ground when you launch the application.
To execute a locally built docker image,
# IMPORTANT!!
# allow connections to X server
xhost +local:root
# NOTE: The realsense camera should be connected to the host.
docker run \
--gpus=all \
--net=host \
--privileged \
--device=/dev/usb \
--device=/dev/media0 \
--device=/dev/media1 \
--device=/dev/video0 \
--device=/dev/video1 \
--device=/dev/video2 \
--device=/dev/video3 \
--device=/dev/video4 \
--device=/dev/video5 \
--env="DISPLAY" \
--env="QT_X11_NO_MITSHM=1" \
--volume=/tmp/.X11-unix:/tmp/.X11-unix:rw \
-it \
distance_assistant/prototype
You might see errors like this:
[ INFO] [1592853388.649425112]: Using transport "raw"
No protocol specified
Unable to init server: Could not connect: Connection refused
(vis_img:105): Gtk-WARNING **: 19:16:28.652: cannot open display: :1
[image_view-5] process has died [pid 105, exit code -11
These errors mean that Docker does not have access to the X11 display server. To fix this issue, ensure that you run the following command before starting the Docker container via docker run
:
xhost +local:root
To turn the current host into a Distance Assistant Kiosk, one can take advantage of the included ansible files.
The kiosking process will alter the existing system to be a dedicated purpose device for running distance assistant, it will boot straight into a limited purpose window manager and run distance assistant in the background.
It is not recommended to do this with a general use laptop or desktop.
Requirements: Ubuntu 18.04.5 or higher on an x86/64 system matching the above hardware specifications.
Go into BIOS, disable secure boot, disable trusted computing
Download an Ubuntu installation image, e.g. https://releases.ubuntu.com/18.04/ubuntu-18.04.5-desktop-amd64.iso
Make sure that this is 18.04.5 or later; earlier versions likely won’t work
Follow this procedure:
https://ubuntu.com/tutorials/create-a-usb-stick-on-ubuntu#1-overview
-
Boot the image ⋅⋅⋅May need to follow prompts on bios to select the USB disk..
-
Select install Ubuntu
-
Select English (US) keyboard layout
-
Select “Minimal installation”
-
Check download updates while installing
-
Check install third-party software for graphics and wifi hardware
-
If you see a checkbox for secure boot, reboot and go back to Configure BIOS
-
Select “Erase Disk and Install Ubuntu” (no LVM/Encryption)
-
Select appropriate timezone
-
Provide User Credentials
- Your name: DistanceAssistant
- Your Computer’s Name: dakiosk
- Pick a username: provision
- Choose a password: changeme
- Confirm your password: changeme
-
Let system install and restart
- log in as provision/changeme
- Pull up a terminal window (Ctrl+Alt+T); or waffle, search for terminal
Download or copy the data from this github repository onto the device.
- Install git:
sudo apt-get install git
-
Get the repository URL. In the upper right of this github repository, click "code", and copy the download URL
-
Clone the repository:
git clone <paste link copied from right pane here>
- Change directory into the repository
cd distance-assistant
sudo apt-add-repository --yes --update ppa:ansible/ansible
sudo apt-get -y install ansible
-
Edit the file distance_assistant_ansible/ansible/group_vars/all
-
Change:
-
da_container_id: set this to the docker tag of the container to use. This defaults to using local image built by the ansible playbooks or by following the Build Instructions
Hosting your container image on Dockerhub or AWS ECR will avoid having to build the container on each host.
-
dnsname_for_healthcheck: set this to the dnsname distance assistant should resolve to confirm that it had basic internet connectivity. This defaults to "amazon.com"
-
set enable_sound to False if you do not want sound
-
Users have reported better stability with the 5.3.x kernels. The first step is using ansible to downgrade the kernel.
This must be executed in the top level directory of this package.
sudo ansible-playbook -i "localhost," \
--extra-vars "base_dir=`pwd`" \
./distance_assistant_ansible/ansible/all.yml \
--tags set_53_kernel
Once the playbook completes execution, the system will reboot.
After the system comes back up, log in again, cd into the distance-assistant directory and run the rest of the kiosking playbooks.
This must be executed in the top level directory of this package.
sudo ansible-playbook -i "localhost," \
--extra-vars "base_dir=`pwd`" \
./distance_assistant_ansible/ansible/all.yml
When complete, the system should reboot to a blue screen and then start distance assistant. If you configured a remote repository, the station may need to download the image which will take some time.
Some devices have both Intel and NVidia GPUs, try adjusting which one is being used:
sudo prime-select intel
Then reboot
This is mostly mitigated by having DA start with a small splash icon. If that doesn't fix it, you can force a screen resize by running a short lived xwindows app. Press "Left_ALT" and "x" at the same time and type "flash_logo.sh", that should display the splash screen again for 30 seconds, then the screen will correctly resize.
If the nvidia drivers weren't installed/enabled there may be an error building the container for DA. Reboot your machine and rerun ansible.
Edit group_vars (per above), and change enable_sound to True or False depending upon if you want to enable or disable sound, then rerun ansible:
Note: you need to be logged in (ssh/etc) as the user with which you provisioned the device and/or a user with sudo rights.
sudo ansible-playbook -i "localhost," \
--extra-vars "base_dir=`pwd`" \
./distance_assistant_ansible/ansible/all.yml
You can also disable/enable sound temporarily (will reset after reboot) by the following commands:
Get Status:
sudo service da_sound status
Stop Sound:
sudo service da_sound stop
Start Sound:
sudo service_dad_sound start