RoboAssistant

Project Description

Our robotics final project will be split into three parts. These three parts consist of a robotic arm, navigation using SLAM, and vision processing. Our task for this final project will be to go to the FIZ, pick up a bin of tools and bring back that bin to Dana 3. We will be using SLAM to navigate to the FIZ, in the FIZ and back to Dana 3. Our robotic arm will be attached to the turtlebot. This arm will be capable of picking up a custom bin and holding until we reach our drop off location. Vision processing will be used to determine the correct bin that needs to be picked up.

The levels of deliverables are as follows:

Level 1: Using SLAM to map IRLL and fiz Using SLAM to navigate to and from fiz Human puts tool in bin on top of rover Obstacle avoidance/dynamic path planning

Level 2: Incorporate arm Vision processing for finding a bin Vision processing for picking up bin

Level 3: Turtle bot knows how to self charge Face Recognition (human robot interaction with only certain humans aka fiz workers) Vision processing for locating specific bin

Instructions for sending robot to a point in the map

Install the following dependencies

$ sudo apt-get install ros-indigo-move-base-msgs

Instructions to run test it:

A map is required to be already saved on the turtlebot station Connect the turtlebot to the workstation Then ssh into the turtlebot and type the following command:

$ roslaunch turtlebot_bringup minimal.launch

in another terminal (ssh into the turtlebot)

$ roslaunch turtlebot_navigation amcl_demo.launch map_file:=<location of the map in the turtlebot.yaml>

on the workstation open a terminal and type the following:

$ roslaunch turtlebot_rviz_launchers view_navigation.launch --screen

Select "2D pose estimate" and select the direction of the of the turtlebot

Then we are ready to go :

edit the code by giving in the set coordinates by using "publish point" from rviz

then in a new terminal:

$ python /go_to_specific_point_on_map.py

Instructions for using AR Tags

NOTE: We assume you have already installed and setup the Turtlebot. Instructions taken from this blog.

Install AR_Track_Alvar in the Turtlebot Laptop

$ sudo apt-get install ros-indigo-ar-track-alvar

Create a launch file

$ roscd turtlebot_bringup
$ cd launch
$ sudo nano alvar.launch

Copy and paste the following code to alvar.launch file

<launch>
    <!--If you are not using gmapping, UNCOMMENT the following line.
        However, if you are using gmapping, you need to run amcl_demo.launch before
        running alvar.launch-->
    <!--<include file="$(find turtlebot_bringup)/launch/3dsensor.launch"/>-->

    <arg name="marker_size"          default="5.0" />
    <arg name="max_new_marker_error" default="0.05" />
    <arg name="max_track_error"      default="0.05" />
    <arg name="cam_image_topic"      default="/camera/depth_registered/points" />
    <arg name="cam_info_topic"       default="/camera/rgb/camera_info" />
    <arg name="output_frame"         default="/camera_rgb_optical_frame" />

    <node name="ar_track_alvar" pkg="ar_track_alvar" type="individualMarkers" respawn="false" output="screen" args="$(arg marker_size) $(arg max_new_marker_error) $(arg max_track_error) $(arg cam_image_topic) $(arg cam_info_topic) $(arg output_frame)" />
</launch>

You can then use this launch file to detect AR tags by using the command

$ roslaunch turtlebot_bringup alvar.launch

To check the output, use the following command. Place an AR tag in front of the kinect camera and it should detect the Tag ID and it's x, y, z

$ rostopic echo /ar_pose_marker

To create AR Tags, use the following command. It'll prompt for the AR Tag ID you want to create and it will create a .png file on the current directory you are running the command. Just print it!

$ rosrun ar_track_alvar createMarker

Instructions for auto-docking

Docking station must be 3 meters in line of sight for the Turtlebot to successfully dock. We assume that minimal.launch is already running. Instructions from here.

On the workstation run:

$ roslaunch kobuki_auto_docking minimal.launch --screen

On a new terminal run:

$ roslaunch kobuki_auto_docking activate.launch --screen

Instructions for sound_playing

On the turtlebot run:

$ rosrun sound_play soundplay_node.py

Instructions for robot_arm_motion

Install rosserial_arduino on turtlebot:

$ sudo apt-get install ros-indigo-rosserial-arduino`
$ sudo apt-get install ros-indigo-rosserial`

On the turtlebot, run the rosserial client application to start the communication with ROS:

$ rosrun rosserial_python serial_node.py /dev/ttyACM0