Solving maze using several state-of-the-art (SOTA) maze-solving approaches: Breadth-first search, Depth-first search, Dijkstra's algorithm, A* search algorithm and Wall-follower algorithm.

I uploaded my work from my internship on: Feb 7, 2023.

• Ubuntu 20.04.2LTS 64bit
• ROS-Noetic
• Gazebo 11
• Python 3

1. Clone this repository

git clone https://github.com/joewong00/maze_solver.git

2. Create catkin workspace

mkdir -p catkin_ws/src

3. Link the respective package to your workspace

sudo ln -fs /path_to_this_repo ~/catkin_ws/src

4. ROS Setup

cd ~/catkin_ws
catkin_make
source ~/catkin_ws/devel/setup.bash

5. Make main python file executable

chmod a+x ~/catkin_ws/src/maze_solver/scripts/main.py

6. Launch python file package for different mazes, change the map config in main.py accordingly (map1.yaml)

roslaunch maze_solver maze1.launch

Maze-Solving in ROS

Create and edit the maze in gazebo, save the maze model in path and change accordingly in world and launch files. To create map, one can use Gazebo plugin like the one in https://github.com/marinaKollmitz/gazebo_ros_2Dmap_plugin.git or any other mapping methods. If the map is not too large, one can hard code the grid value (1 for wall and 0 for passage), convert into pgm and create the map yaml file.