RAS Docker Walkthrough

This document provides a comprehensive walkthrough of the RAS Docker environment structure and functionality. Use this guide to navigate the codebase and understand how to use the system.

Directory Structure

.
├── apps                   # Contains application-specific code
│   ├── ras_robot_app      # Robot controller application
│   └── ras_server_app     # Server application for remote control and monitoring
├── assets                 # Static assets (images, 3D models, etc.)
├── configs                # Configuration files for the system
│   ├── experiments        # Robot experiment definitions
│   ├── lab_setup.yaml     # Laboratory environment configuration
│   ├── objects.yaml       # Object definitions for manipulation
│   └── ras_conf.yaml      # Main RAS system configuration
├── context                # Docker context files
├── repos                  # External repositories and dependencies
├── ros2_pkgs              # ROS2 packages for robot control
│   ├── ras_bt_framework   # Behavior Tree framework for task execution
│   ├── ras_core_pkgs      # Core ROS2 packages for robot functionality
│   └── ras_sim            # Simulation environment packages
├── scripts                # Utility scripts for various tasks
└── env.sh                 # Environment setup script

---

Prerequisites

Before starting, ensure your system has the following installed:

1. Ubuntu OS (RAS officially supports Ubuntu)

2. Git (To download the required files)

   sudo apt install git

3. Docker (For running applications in containers)

   sudo apt install docker.io

   or

   Docker

   Verify Docker installation:

   • run the following command to check for the docker installation:

   docker

4. Nvidia Container Toolkit (If your system have nvidia GPU installed)

   Check for Nvidia GPU,

   nvidia-smi

   If not installed, it will show versions to install the nvidia drivers.

   If already have then run the following command to install nvidia container toolkit.

   sudo apt-get install -y nvidia-container-toolkit

5. vcstool (For managing repositories)

   python3 -m pip install vcstool

   If pip is not installed:

   sudo apt install python3-pip

6. Argcomplete (For auto-completing commands)

   sudo apt install python3-argcomplete

7. Stable Internet Connection (Required to pull Docker images properly)

8. xArm Robotic Manipulator (With Dedicated Desktop Setup for Robot Setup)

9. Robot Hardware Connection (For Robot Desktop Setup) Use an Ethernet cable to connect the robot and your PC. Default IP of xArm: 192.168.1.111

Set Manual Ethernet IP Address:

1. Go to Settings → Network → Ethernet → IPv4
2. Select Manual and enter:
   • IP Address: 192.168.1.78 (or any IP in the xArm's subnet)
   • Subnet Mask: 255.255.255.0

Check connection:

ping 192.168.1.111

If packets are sent and received successfully, the connection is established.

---

Installation

Step 1: Clone the Repository

Download the RAS Docker workspace files by running:

git clone --recursive https://github.com/ras-ros2/ras_docker

---

Step 2: Set Up Environment

Go inside the ras_docker folder and set up the environment:

cd ras_docker

Source the environment using the following command:

. env.sh

---

Step 3: Check Available Commands

List available commands for the RAS Docker Interface (RDI):

ras -h

---

Step 4: Initialize the App

Set up the application:

ras <app> init

Note: If the internet is unstable, this command may build the image locally. Run the following command to force pull the image from DockerHub:

ras <app> init -i

This creates the ras_server_app/ras_robot_app folder inside the apps directory.

---

Step 5: Build the Application

Build the application:

ras <app> build

This command will build

• the necessary Docker image for the server.
• the ROS 2 workspace inside the ras_server_app/ros2_ws and ras_robot_app/ros2_ws directory. If the ROS 2 workspace is not built then run it again.

Note: If you've downloaded new images or updated an existing one, make sure to clean the build before rebuilding:

ras <app> build --clean

---

Step 6: Configure the Network

Before running the , configure ras_conf.yaml:

nano ras_docker/configs/ras_conf.yaml

Update transport settings based on your network setup:

Default:

Running on the Same Machine (Localhost)

ras:
  transport:
    implementation: default
    file_server:
      use_external: false
      ip: localhost
      port: 2122
    mqtt:
      use_external: false
      ip: localhost
      port: 2383

For more examples, see Network Configuration Setup README.

---

Step 7: Run the App

Start the app inside a Docker container:

ras <app> run

For more details, see Applications README.

---

Step 8: Access the Container

To log into the running container:

ras <app> dev

To exit:

exit

or press Ctrl + D.

---

Step 9: Kill the App

To stop the app, use:

ras_kill

Done!

Your app is now set up and running. If you face any issues:

• Follow the troubleshooting steps provided.
• Restart your system.
• For docker troubleshooting, Docker Network and Configuration README.
• For GUI interaction, UFactory Studio README.
• Refer this RAS Server Setup README. for complete Server Setup.
• Refer this RAS Robot Setup README. for complete Server Setup.
• Refer this RAS Workspace Setup README. for complete Physical workspace Setup.

For further assistance, refer to the documentation or raise an issue on the project repository.