TGIF

export DISPLAY=$(cat /etc/resolv.conf | grep nameserver | awk '{print $2}'):0

docker run --name open_rmf --gpus all -it -e DISPLAY=$DISPLAY osrf/ros:humble-desktop

New-Item -Path 'D:\temp\Test Folder' -ItemType Directory

apt install ros-humble-rmf-*

sudo apt-get install python3-pip

set-variable -name DISPLAY -value 150.165.167.44:0.0

docker build -t open_rmf_demos .

docker run -it -e DISPLAY=$DISPLAY --gpus all --name open_rmf_running open_rmf_demos

# Step 4: Install rest of OpenRMF binaires via rosdep
CMD cd ~/rmf_demos_workspace
CMD rosdep install --from-paths src --ignore-src --rosdistro humble -y

# Step 5: Build OpenRMF Demos workspace
CMD source /opt/ros/humble/setup.bash
CMD colcon build

Open a docker terminal

Get the image running as a container in the background: docker run -d -it <image_id>

Tip: docker ps will show the container_id that you just fired up from said image.
Per @eltonStoneman's advice: docker exec -it <container_id> bash
Now your docker terminal is showing an interactive terminal to the container.
Open up another terminal and perform step 3 to create another interactive terminal to the container. (Rinse and Repeat)

https://docs.nvidia.com/ai-enterprise/deployment-guide-vmware/0.1.0/docker.html

https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/user-guide.html#gpu-enumeration

https://www.youtube.com/watch?v=cMsIT2otEjA