How to build a maze

To: micromouse@dcs.rhbnc.ac.uk
Subject: How to build a maze
From: A Johnstone <adrian@dcs.rhbnc.ac.uk>
Date: Thu, 8 Jun 2000 12:23:38 +0100 (BST)

This week's topic is... how best to build a maze?

The micromouse mazes that I have seen are serious undertakings. Usually
they are in four parts with removable walls. Each part is 4-5 feet on an
edge with strengthening underneath and transport is not much fun. Alan
Dibley has some views on simpler wall systems, but the mazes that I have
seen have central posts with slots in then into which the walls are
fitted.

The UEL maze has an MDF top painted with blackboard paint; timber
underpinnings that are aligned using plug-and-socket wooden tongues plus
coach bolts; about 300 wooden posts each with a dowel underneath and a
similar number of walls cut from melamine covered chipboard, painted red
along one edge and with a routed tongue at each end. They paid about 1K
pounds to have it made, and the constructor swore never again... 

I have a fantasy maze but no way of making it. The maze would be made up
of smaller floor parts, possibly the size of a 2x2 or even an individual
maze cell. The floor part would be made of injection moulded plastic with
the strength and durability of Lego parts. It would have a nice matt black
finish which was a little grippy, like the tops of the keys on most
keyboards. The floor parts would snap together using a robust tongue and
groove arrangement that ensured a tight fit and structural rigidity. The
posts and walls would similarly be injection moulded plastic and would
snap into holes in the base parts.

Ah, if only we could persuade Lego to divert one of their factories to our
ends for a few minutes (oh and make the moulds, of course). Each maze
would need about 300 floor cells, and a production run of a few thousand
would enable us to have two mazes (see below) and kits of 5x5 sized mazes
to be sold off to mouse builders.

Why two mazes? Well, we should aim to make micromouse more of a spectator
event in future years. Having two mazes with simultaneous runs would be
great, especially where we have a pair of similar mice such as Ken's. 
Even in the normal competition it would be fun to see a direct race, but
my proposed sprint event would particularly benefit.


For me, this is not just a theoretical problem as I've got to get on and
make a Royal Holloway maze for next year, so all advice as to materials
and construction methods will be gratefully received.

Don't forget to send replies to micromouse@dcs.rhbnc.ac.uk if they are for
public consumption.

                         Adrian



Dr Adrian Johnstone, Senior Lecturer in Computing, Computer Science Dep, 
Royal Holloway, University of London,  Egham, Surrey, TW20 0EX, England.
Email a.johnstone@rhbnc.ac.uk Tel:+44(0)1784 443425 Fax:+44(0)1784 439786