RoboND-Project5

Implementation of project 5, Home-service Robot project Robotics Software Engineering Nanodegree by Udacity

Introduction

This repository includes two similar implementation of the "Homeservice Robot project". One is using the turtlebot, the other is using a custom-build robot called my_robot. Currently these two robots can map the gazebo environment with teleop_control or naive automatic exploration, and can navigate inside the different goals inside the environment. Additionally, my_robot also can automatically explore and map the environmenr. Later the function for my_robot may be updated, equipping it to catch and transfer objects.

automatic exploration start

automatic exploration end

myrobot navigate to pick up zone

myrobot navigate to drop off zone

Installation

This project is developed with Ubuntu 16.04 + ROS-kinetic Download this project into your workspace, and download the kinetic version of the following official packages

• gmapping
• turtlebot_teleop
• turtlebot_rviz_launchers
• turtlebot_gazebo

Also install xterm with

sudo apt-get install xterm

Then please follow the instructions in the next section to modify the path in several shell scipts, otherwise those scipts won't work properly.

Path Issue in script files

In test_slam.sh, a file path is included in line 2; In test_navigation.sh, home_service.sh, pick_objects.sh,file paths are included in line 2 and line 4. Please make sure to change the paths from

"/home/centuryliu/robotic_self_learning_ws/src/my_robot/worlds/open_classroom.world"

to

"(path_to_your_workspace)/src/my_robot/worlds/open_classroom.world"

Otherwise the scripts won't work properly

Commands

After fixing the path issue, first go to the /map_my_environment/scripts directory, open a terminal in this directory. Give permission for the automatic_mapping.py

chmod +x automatic_mapping.py

Then go to the scripts directory in your workspace, open a terminal in this directory. Give permissions for all the shell scripts to be executable

chmod +x *

Then in the same terminal enter

./home_service.sh

to see the behavior of turtlebot; or use

./home_service_my_robot.sh

for my_robot

To automatically map the environment, use

./slam_auto.sh

File structure

├──add_markers                    # add_markers C++ node
│   ├── src
    │   ├── add_markers.cpp        # add markers for home_service
    │   ├── add_markers_only.cpp   # add marker at pick-up for 5s, pause 5s, then add marker at goal
│   ├── rviz/nav_with_marker.rviz  # view_navigation.rviz + marker display
|   ├── launch
    │   ├── add_markers.launch  # launch the add_markers node and display in rviz
    │   ├── add_markers_only.launch  # launch the add_markers_only node and display in rviz
├──locate_my_robot                 # package for localizing my_robot
│   ├── launch/amcl_only.launch    # launch file for using amcl for my_robot with slam map
│   ├── map
    │   ├──slam_map.pgm            # slam map pgm
    │   ├──slam_map.yaml           # slam map yaml. The map is created by using slam_gmapping, coordinates of the map are in line with gazebo
    │   ├──open_classroom_map.zip  # high resolution map created by using pgm_map_creator. Coordinate rotated pi/2 with respect to gazebo
│   ├── scripts
    │   ├──my_robot_teleop.py      # teleop my_robot
    │   ├──self_rescue.py          # naive exploration implementation of my_robot, used in slam_auto_naive.sh
├──map_my_environment              # map the environment
│   ├── launch/my_robot_mapping.launch # call slam_gmapping, args configured for my_robot
│   ├── maps
    │   ├──auto_built_map.pgm      # a copy of the pgm file of the auto-created map
    │   ├──auto_built_map.yaml     # a copy of the yaml file of the auto-created map
│   ├── scripts
    │   ├──automatic_mapping.py    # ros node for automatic mapping. Plan a new target for exploration at each iteration, and send the target to movebase
    │   ├──map_listener.py         # class for getting and pre-processing the map published by slam-gmapping
    │   ├──map_split_detection.py  # utility function used when planning target
    │   ├──odom_listener.py        # class for getting the odom of the robot
    
├──my_robot                        # package containing simulation environment and robot description
│   ├── launch/final_project_world.launch # launch the world for this project and the my_robot
│   ├── worlds/open_classroom.world # world created for my
├──pick_objects                    # pick_objects C++ node
│   ├── launch/use_move_base.launch # launch file with args specified for my_robot to use move_base
│   ├── src/pick_objects.cpp       # the pick_objects node
├── scripts                     # shell scripts files
│   ├── add_marker.sh              # show marker to represent pick up and drop off
│   ├── add_marker_my_robot.sh     # same for add_marker; for my_robot, same for all scripts with "_my_robot"
│   ├── home_service.sh            # the main script for project
│   ├── home_service_my_robot.sh
│   ├── pick_objects.sh            # navigate turtlebot to pick up zone and drop off zone
│   ├── pick_objects_my_robot.sh   
|   ├── slam_auto.sh               # automatically map the environment
│   ├── slam_auto_naive.sh         # naive implementation of automatic mapping using my_robot
│   ├── slam_my_robot.sh           # teleop my_robot to map the environment
│   ├── test_navigation.sh         # test navigation configuration of turtlebot    
│   ├── test_slam.sh               # test slam configuration of turtlebot
└──