MazeSolver

Abstract

This project entails designing and constructing an autonomous robot, MazeSolver, to solve mazes using the Flood Fill algorithm. Motivated by the rising significance of autonomous navigation in disaster response, smart manufacturing, and logistics automation, the objective is to demonstrate real-time decision-making through sensor feedback and algorithmic path planning [3, 4, 5]. The robot employs three IR sensors and a VL53L0X Time-of-Flight LiDAR for environmental mapping, dynamically updating its internal map upon detecting obstacles [15, 16]. Utilizing the Flood Fill algorithm, it adapts its path in real-time, striving to select the shortest route despite an initially unknown maze layout. Development involved hardware design with an ESP32 microcontroller, sensor calibration, programming in Arduino C++, and testing in a physical 7x6 maze, with movement controlled by two DC motors with encoders via an L298N driver and orientation maintained by an MPU6050 gyroscope [9, 12]. Communi- cation is facilitated by a remote unit with a second ESP32 using the ESP-NOW protocol [1]. MazeSolver distinguishes itself from predecessors like Micromouse through its real-time wall detection and dynamic replanning, enhancing adaptability for applications such as warehouse robots, search-and-rescue units, and agricultural machines in unpredictable environments.