How Big is a Smart Micromouse?
The micromouse rules allow for a lot of latitude in designing a robot. The cell size is 180 x 180 mm. with 12 mm corner posts, which means that the width of the path within the maze is a maximum of 168 mm. Of course, the mouse has to be considerably thinner than the maximum from a practical standpoint. The rules also allow for the mouse to extend above the maze walls to the point that it could occupy a space as large as 250 x 250 mm.
One design approach is to build the mouse so that it has sensors that read the tops of the walls in much the same way that a human would perceive it looking down on the maze from above.
Note: Photos taken at the 25th All Japan Micromouse Competition in Tokyo, Japan - November, 2004
This approach is certainly workable and has achieved a level of success. It simplifies the wall detection problem, and has the added advantage of being able to peek over the walls of the current cell to identify connecting walls in adjacent cells. The most obvious disadvantage is the high center of gravity dictated by the mouse height. At low speeds while mapping the maze this is not a significant problem. During the speed runs however, the center of gravity is likely to play a role in determining the overall mouse stability and may limit the top speed it can achieve reliably.
Another viable design approach is to keep the mouse below the level of the walls, totally within the maze as if it was a real mouse. Robots designed from this perspective seem to fall into several broad classifications.
The "PC Stack"
Boards, batteries, and motors are stacked in vertical layers.
The "Flying PC Board"
Extremely low center of gravity - well below the wheel axis - with excellent aerodynamics.
A minimalist design that takes the previous example to it's logical extreme.
And, the "Race Car"-
My personal favorite - at least at this point in time....
The top competitors have structured their designs so that the mouse can identify and travel along diagonal paths in the maze. While this enables them to shorten the total path while eliminating the need for time consuming turns, it also constrains their width.
In order to travel on a diagonal, the mouse has to have sufficient clearance, not only to fit through the diagonal space, but also to allow for drifts from calculated positions. For example, coming off of a high speed straight away and making a turn into a diagonal leg is likely to cause the rear end of the mouse to overshoot, or the entire mouse body may drift towards one side or the other of the intended path.
The maze post centerlines are laid out so that the diagonal gap between them is just over 127 mm - without taking the width of the posts themselves into consideration. If a design provides for an over generous clearance safety margin, then it will give up precious hundreds of a second. On the other hand, if it scrimps on safety margin it runs a significant risk of impact. During the preliminary competition on Saturday, several of the mice ran into this post in the maze:
Some of them clipped the post at a diagonal, while others ran straight into it. The problem may have been tied to their sensors misreading the post as the edge of the open space since it was slightly discolored when compared to the walls and other posts, or there may have been some other factors at play. In any case, I want to keep the design safety margins as generous as possible in the beginning, then take more risks later in the game. It may prove beneficial to design the control program with switchable parameters that allow for taking more risks on successive speed runs.
Unfortunately, a commitment has to be made to the overall design approach and the corresponding dimensions early in the process. With that in mind, I'm going to constrain the width of my micromouse design, tentatively named "Peevee", to a maximum of 65 mm. I may have to change that later, but for now I think it's a workable goal.
You might also enjoy: