Your micromouse chassis holds all the bits together. It must provide support for motors, batteries, sensors and the controller. ideally, it will be light, strong, small and easy to construct.
The materials you use will probably depend on what is availalable and your existing skills. Appearance is probably not going to be your main concern so don’t worry about wires wandering about and blobs of glue in clear view. Performance is the main aim, everything else is secondary. Naturally, if you can have adequate performance and look gorgeous then go for it.
Decisions have to be made about sensor types and layout; motor types and their layout; how you will steer; where the batteries will go; how the controller will be placed; how everything gets fixed together; how you are going to rebuild if things turn out wrong and – by no means least – how big it should be.
There are nearly as many solutions to these problems as there are mouse builders. In spite of that, there seem to be relatively few really innovative solutions out there at the moment.
Planning your chassis is probably something to do pretty early. If you are a programming type, it might be tempting to work on emulation and code fragments. Well, until you have something reliable to run around in, there is little point in knowing how to drive is there?
In general, you can write software to deal with and, perhaps, overcome mechanical limitations. These are much better dealt with in design rather than after construction. For example, excessive backlash in the gears is not necessarily adequately solvable in software. You can, however, compensate for a slight difference in wheel diameter. In any case, you are going to need a lot of actual testing. For that you need a mouse. Design and build early and do your software development on it rather than a simulation. Which will actually be chasing around the maze on the day – your micromouse or your PC?