RRRobot!

Page Links

• Home
• Documentation
• GEAR

Table of Contents

• Contributors
• Documents
• Introduction
• Development Guide
  • Docker Compose Services
  • GUI Support
  • Running Docker Containers
  • Gazebo Grasping Simulation
  • Tips & Tricks
• Source Code Documentation
• GEAR Final Simulation

Contributors

Project Team

• Sravan Balaji (balajsra@umich.edu)
• Chenxi Gu (chenxgu@umich.edu)
• Jake Johnson (thejakej@umich.edu)
• Derek Witcpalek (dwitcpa@umich.edu)

EECS 467 W20 Course Staff

• Prof. Chad Jenkins (ocj@umich.edu)
• Xiaotong Chen (cxt@umich.edu)
• Jana Pavlasek (pavlasek@umich.edu)

Documents

1. Project Poster
2. Project Proposal
3. Final Project Report (PDF) (Overleaf)
4. Final Video

Introduction

Repository for UMICH EECS 467: Autonomous Robotics (W20) RRRobot! project.

To get started, you will need to install Docker on your system. Information on what Docker is, how to install it, and how to use it can be found in Getting Started with Docker.

Once installed, you can run our final simulation by following the instructions on the GEAR page.

Information about our source code can be found on the Documentation page.

Development Guide

At this point, you should have Docker Desktop, Toolbox, or Engine setup on your host system. In this section, you will find a list of the different docker-compose services available and how to run them.

Docker Compose Services

ros-dev

ros-dev is the development container for ROS. This is used to provide all the necessary dependencies for working through the ROS tutorials and creating ROS packages.

ros-turtlesim

ros-turtlesim is a container that tests whether GUI support is working. This will install the ros tutorial packages and run the turtlesim GUI with keyboard control. If X forwarding is setup correctly, you should see a window open up with a turtle on a blue background. Use the arrow keys in the container terminal to move it around and press q to quit.

ros

ros is a container based on the ROS Melodic docker image. This is provided for convenience, but not used for running the final simulation.

gazebo

gazebo is a container based on the Gazebo Server 9 docker image. This is provided for convenience, but not used for running the final simulation.

gear

The GEAR container includes the Gazebo Environment for Agile Robotics from the Agile Robotics for Industrial Automation Competition 2019. This is used for running the final simulation.

GUI Support

If using VcXsrv for Windows to enable GUI applications, run XLaunch using config.xlaunch. This will enable the following settings:

• Display Settings
  • Multiple Windows
  • Display Number: -1
• Client Startup
  • Start No Client
• Extra Settings
  • Clipboard
    • Primary Selection
  • Native OpenGL
  • Disable Access Control

The main thing to note here is that Native OpenGL is disabled and Disable Access Control is enabled. When running Gazebo GUI applications through a Docker Container in WSL, enabling Native OpenGL resulted in the following errors:

libGL error: No matching fbConfigs or visuals found
libGL error: failed to load driver: swrast
libGL error: No matching fbConfigs or visuals found
libGL error: failed to load driver: swrast

As mentioned in No libGL libraries when running Gazebo from ROS, this can be resolved by disabling No native OpenGL in the VcXsrv configuration.

You will need to update your environment variable with your IP Address to enable X forwarding. You can create a file called .env in the same directory as the docker-compose.yml file with your IP address as shown below. You can also edit the docker-compose.yml to remove the dependence on the IP_ADDRESS environment variable.

# .env for docker-compose
IP_ADDRESS=<IP Address Here>

Running Docker Containers

1. Start Docker Machine (default)
   • WSL: docker-machine.exe start default
   • Other: docker-machine start default
2. Set Docker Machine Environment
   • WSL: docker-machine.exe env
   • Other: docker-machine env
3. Navigate to src folder
   • cd /PATH/TO/rrrobot/src
4. Use Docker Compose to run a service (refer to docker-compose.yml or Docker Compose Services)
   • docker-compose run --rm <service_name>

Gazebo Grasping Simulation

1. Source ROS Setup
   • source /opt/ros/melodic/setup.bash
2. Build the drivers for the simulation
   • cd /app/rrrobot_src/src/simulation_drivers/
   • source build.sh
3. Start ros master node
   • roscore &
4. Run the gazebo simulator - this will bring up gazebo with a robotic arm
   • gazebo /app/rrrobot_src/world/rrrobot.world
5. Run control and perception programs

Tips & Tricks

• Clear space on docker machine
  • docker system prune --volumes
• See running containers
  • docker ps
• Attach a new terminal to a running container
  • docker exec -it -u root <container> bash

Source Code Documentation

Documentation on our source code can be found on the Documentation page.

GEAR Final Simulation

Instructions for running our final GEAR simulation can be found on the GEAR page.