Micromouse — the Little Robot That Solves a Maze

Micromouse is one of the oldest robotics competitions in the world and still one of the purest. The rule is elegant: a square maze, a small robot, and a goal at the centre. The robot must explore the maze itself, then run the fastest route.

Two phases

Exploration run

The robot starts from a corner and searches the maze. Speed does not matter here; building a correct wall map and reaching the centre does.

Speed run

With a map in hand, the robot runs the shortest — really the fastest — path at maximum acceleration. The split between these two phases is the essence of Micromouse: understand first, then execute.

A subtle point: the shortest path is not always the fastest. A route ten cells longer but with no turns usually beats a shorter one with five 90-degree corners, because every corner costs speed. Serious robots optimise for time, not distance.

The flood fill algorithm

The classic Micromouse solution: give every cell a number for its distance to the centre. Centre is zero, its neighbours one, theirs two, and so on. The robot then simply always moves to a neighbouring cell with a lower number than its own.

The beauty is in the update: each time the robot discovers a new wall, the numbers are recomputed and the optimal path corrects itself. The same algorithm appears in Rescue Maze.

Hardware challenges

Moving exactly one cell: a small rotation error becomes a large drift after ten cells. Encoders and a gyroscope are essential.
Accurate 90-degree turns without wheel slip.
Continuous wall correction: the robot keeps itself centred in the corridor using side-wall ranging.
Acceleration and braking: in the speed run, acceleration management separates first place from tenth.
Size: everything must fit in a very small footprint.

Why it teaches so well

Because it demands algorithms (exploration and planning) and precision engineering (motion control) simultaneously, and neither alone is enough. Great algorithms with poor motion control produce a wrong map; great mechanics with weak algorithms never find the optimal route.

Where to start

First a robot that drives straight down a corridor and turns exactly 90 degrees.
Then simple exploration with a left-hand-on-wall rule.
Then building the map in memory.
Then flood fill and path planning.
Finally, speed optimisation.

If you have never built a robot, start with the line follower league — the basic motion control learned there is exactly the prerequisite here.

Frequently asked questions

How does Micromouse differ from Rescue Maze?

Micromouse is purely maze exploration and speed; Rescue Maze adds victim identification, rescue-kit deployment and ramps.

What knowledge is needed?

Encoder-based motor control, distance sensing and microcontroller programming. For the algorithm, arrays and queues suffice.

What age suits it?

Generally teenage upward, since it needs both mechanical precision and algorithmic understanding. Younger students should start with line following.

Want to compete in Micromouse?

Novin Zehn competition teams train from zero to the world stage — in Tehran and live online.