Micromouse Robot System Using the Flood-Fill and Backtracking Algorithm

Abstract

A micromouse robot is an autonomous robot designed to navigate mazes and find the shortest path to a destination point. This research aims to design a micromouse robot software system using a flood-fill algorithm that can move without touching the walls and determine the direction and angle of rotation when encountering a dead end. The system uses an Arduino Nano as a microcontroller, infrared sensors for wall detection, DC motors with rotary encoders, and a 2-channel motor driver, with a maximum robot dimension of 16.8 cm × 16.8 cm. The software implementation uses a Breadth-First Search (BFS)-based flood-fill algorithm for optimal path mapping, a 6-state state machine to regulate execution flow, Proportional-Derivative (PD) control with parameters Kp = 2.0 and Kd = 0.05 for movement stability, and trapezoidal velocity profiling for smooth movement. Test results show that the robot can complete a 4×4 maze in an average of 45 seconds, with position accuracy of ±5 mm and an angle deviation of less than 2 degrees. The PD control system provides a responsiveness of 32 ms, and velocity profiling reduces wheel slip by up to 90% and increases energy efficiency by 25%. This research contributes to the development of mobile robot technology and robotics learning media.