Team Vegavath · Robotics
Maze Solver
An 8x8 autonomous maze-solving robot by Vega Vath Racing. Pick a start and goal, choose the robot's starting heading, and watch it find the shortest route.
How it works
This page works from the known map of the competition maze. A breadth-first search computes the shortest route between the start and goal you pick, then converts that route into the exact turn-by-turn commands the robot would execute from its starting heading. Because the map is known ahead of time, the route shown here is genuinely the shortest one, not a guess refined over several runs.
The physical robot
| Controller | Arduino Uno |
|---|---|
| Heading / turns | MPU6050 IMU |
| Front sensor | HC-SR04 -- wall detection and safety |
| Side sensors | Left + right HC-SR04 -- wall sensing and corridor centering |
| Distance | Right-motor encoder |
| Motor driver | TB6612FNG |
The robot does not blindly trust motion timing. Even while executing a route it already knows from the map, it keeps verifying its surroundings with its sensors, correcting its heading with the IMU and its distance with the encoder rather than assuming a fixed number of milliseconds equals one cell.
Planned autonomous mode
A second mode is planned separately: instead of being handed the map, the robot would build and maintain its own maze state onboard and run flood-fill replanning as it discovers walls, updating its route on the fly. That discovery mode is a distinct piece of work from the known-map shortest-route solver shown on this page.