This was my final project for a robotics course in UMD. (ENPM809E - Python Applications for Robotics)
This algorithm was implemented using the Robot Operating System (ROS) libraries, Gazebo as simulator and Python as programming language.
The goal of this project is to solve a maze problem using only one Turtlebot 3 robot which comes equipped with a laser scanner.
We had to achieve the following tasks:
- Detect a wall
- Follow the wall
- Avoid collision with obstacles
- Find the exit
We had to accomplish the Tasks automatically.
- ROS: Melodic
- Ubuntu: 18.04LTS
- Python: 3.6.9
- (Gazebo: 9.0.0)
To execute the same, I created one python package(FinalProject_ws) which has one module (motion_plan). This module can be used in any catkin space if configured as per the tree below.
.
└── catkin_ws # Your simulation workspace
├── src # Cloned from this directory
│ ├── maze
│ │ ├── launch
│ │ | ├── maze.launch
│ │ ├── world
│ │ | ├── maze.world # Folder that contains maze 1's definitions
│ │ | └── maze2.world # Folder that contains maze 2's definitions
│ │ ├── CMakeLists.txt
│ │ └── pakage.xml
│ ├── motion_plan
│ │ ├── nodes
│ │ | ├── follow_wall.py # Folder that contains motion planning for the robot
│ │ ├── src
│ │ | └── ...
│ │ ├── CMakeLists.txt
│ │ └── pakage.xml
│ └── CMakeLists.txt
├── devel # Folder created after compilation
└── build # Folder created after compilation
This program consists of various fiunctions to check if the bot has detected a wall and based on the distance from obstacle on right side and left side, the program will either enter the algorithm for Right wall follower or the Left wall follower.
-
Download the 'FinalProject_Nimbekar.zip' folder from Canvas and unzip it.
-
Run the
roscore
command in terminal to start ros. -
Run the maze in terminal one:
roslaunch maze maze.launch
- Then run the Code in terminal two:
rosrun motion_plan follow_wall.py
- Manually stop the code once the maze is cleared.