The goal of this project is to implement simultaneous mapping and localization (SLAM) algorithms that allow a robot to explore an unfamiliar maze. Using LiDAR and wheel encoder data, we implemented SLAM using particle filtering (Monte Carlo Localization) for sensor fusion and localization, and A* for path planning and exploration. The robustness of the algorithm was proven by successfully exploring a difficult maze.
The following is a video about how the robot explores the maze:
If you want to skip the video, these are the photo of the true maze and the map that the robot gets after exploring the maze along the red route:
