LAB::Sarthak_Varchasvi

Project

Micromouse Maze Solver

Autonomous maze-solving robot with Sharp IR wall sensing, rotary encoders, magnetometer with tilt compensation for accurate 90° turns, and Flood Fill optimal path algorithm.

Problem Statement

Navigate unknown mazes consistently within competition time limits while handling sensing noise, tight turns, and accumulated heading error across multiple runs.

System Design

Magnetometer with tilt compensation provides absolute heading reference — eliminating encoder-based turn error accumulation after 4–5 turns. Encoders handle straight-line PID. Flood Fill assigns distance values to each cell for optimal path computation.

Key Challenges

Encoder-only turns accumulate heading error — magnetometer with tilt compensation was the key insight. Reliable wall-distance estimation in noisy environments.

Results

1st Place — MicroMouse Challenge, Technex '24, IIT-BHU.