Pure speed is not the primary aim when building a micromouse. On the other
hand, it is a race so, if you want to win, you have to be quicker
than the rest.

It is worth making the point here that you must find the centre. Without
that you might as well just bat about the maze at breakneck speed, hoping
to stumble into the centre by chance. A strategy but not the best.

Unfortunately, just about all your problems get worse as your mouse runs faster. To start with, you are better off with a ‘slow but sure wins the race’ philosophy.

Wen you are confident that your basic algorithms all work and that you can reliably find the centre and the best route for your mouse dynamics, by all means turn up the volume.

A common strategy with competition mice is to map out the maze and find an appropriate maze then run the route faster and faster. It is best to stop before you crash as the consequences of a crash could be pretty severe.

Maximum speed is likely to be limited by one or more of a number of factors:

How far ahead can you see a wall. The further away you can see a wall, the more time you have to stop before hitting it.

How good are your brakes. Related to the previous item good brakes will mean you can afford to detect a wall at a closer distance and still be able to stop. Test away but remember that you will not know how much grip is available until you are actually on the competition maze which may be dry and dusty just where you wish it were not.

Correcting for steering errors. It is a rare mouse that can travel unaided along a corridor. The faster you go, the harder that task becomes. You cannot just increase the corrective action to compensate for speed. That way lies disaster. Just think of how many cars you have seen in ditches where the driver tried just that technique.

Can you see side passages? Slick exploration will allow you to detect turnings and decide, on-the-fly, whether you are better off going straight on or following the turning. This is best done without stopping. Can you recalculate the route, decide and make the appropriate manoeuvre in the time between detecting an opening and having to start the manoeuvre?

How fast is your processing loop. The basic processing loop that runs when your mouse is travelling should not be wasting too much time. This is unlikely but keep processor-bound delays to a minimum. For example, don’t get caught reflooding the maze, updating an LCD display or making some (supposedly) cool sound effect when your mouse should be looking where it is going.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.