Everyone builds robots differently.
There are so many materials and techniques that 'how' you build a robot is a huge topic, but some basic ideas are important to nail down and will help anything work better.
The 'virtual robot' penalty clause
is really there to stop people talking and get them building. It's
much more fun for everyone to have a lots of silly designs that do
simple things than just to have a lot of people talking about
complex machines of mass destruction (and unless you have the mass
you can't really do the destruction).
Some kind of movement is required to demonstrate robotic life. This
can be anything from a push start to an
electric motor. Elastic materials
have a good success rate, and if in doubt pulling something around
from outside the arena with a bit of string will
provide some interest (and even an emergency weapon).
Electric motor with string
steering is a very simple means of control, though
steering is relatively unimportant if the robot is imaginative
enough as spectators and other competitors will turn entertaining
robots round and send them back into the fray. An electric
motor driving wheels is relatively simple to produce, and
fairly reliable. Turning corners can be a lot more difficult
though.
The three methods I have used are:
a) Asymmetric design - using a cam or wheels of different sizes to make the robot turn in a circle of predetermined size.
b) String - simply pull the front of the robot in the direction you now wish it to go.
c ) Silliness - anyone that's laughing at your robot is very likely to turn it round to have another go when it reaches the edge of the arena, themselves!
Of these silliness is the mechanically simplest, but requires inspiration rather than perspiration and is not always possible. Walking machines have also been made, this though needs more imagination - and either better mechanical knowledge or application of arcane technology (such as K'Nex...). Hopping, hovering, wobbling and other methods of locomotion are all possible to achieve and delight the crowds and umpires most, but wheels are simplest.
The biggest problem with many electrical motors is that they tend to run at high speed, this means that either small wheels or some kind of gearing down is required. Many children's toys have such gears in though - including non-electrical ones, such as friction motor baby toys. These friction motors (wind up the flywheel and let them go) can also make perfectly good drives in themselves, but are very difficult to build from scratch and are best scavenged. Charity shops are often a good source of half broken toys for such bits.
The most basic robot is something that moves and either survives in
an entertaining manner, dies in and entertaining manner, or causes
other robots to die in an entertaining manner. Once you have
something that can move you have to work what to be entertaining
at.
I learnt that an entertaining method of movement in a fairly
robust design is all that is really needed. To build something
specifically to die well is a challenge, and arguably more
difficult than either of the others. I have had a number of idea in
this direction but want to try them myself first before inflicting
them on others...
Survival must be entertaining, or you will simply have recreated the horror that is the bimblebot, and will neither win prizes nor (more importantly) make people happy.
An entertaining method of locomotion, and a 'Heath Robinson' style of construction both go down well. Letting people see the bits moving around (and, ideally, slowly falling off) is good. Anything that looks fairly fragile but can take and impact goes down well too - this has been a major advantage of K'Nex, but a lightweight frame work without body panels hiding the internals can be made with many different things - anyone who has played with Balsa wood in their youth is off to a head start here. As soon as you hide the internals away you are producing a more professional looking machine and loose some of the entertainment value (although a really silly bodyshell - especially with a TV media image, based on past experience - will make up for this). Generally speaking the more 'combat worthy' something looks the less entertaining it's mere survival will be...you need to kill things to make a glossy image work...
Destruction is difficult.
As mentioned earlier, for most simple methods you need mass. A robot that is going to knock other robots really needs to weigh enough to produce a lot of traction, or have sufficient combined speed and weight to be able to physically beat up other robots. This has added problems. Getting violent is something that I have seriously considered, but so far avoided due to the extra complication required. Something simple that works is always more interesting than something complex that just sits there...
Basic combat is simply ramming things. This does not require and for of steering - most other robots are unsteered so eventually your robot will hit another.
Remote radio control or similar technology (possibly string...) is of use if you want to hit particular other robots, but radio control takes serious investment...Most cheap controllers use the same wavelength (someone can no doubt give the techie details) and so you really need to buy a proper hobbyists remote control and spare crystals so that you can change frequency if there is a clash. This is likely to cost in the £50-100 pound mark - but the bits will be reusable so the expense will diminish if you are serious about entering regularly.
The most simple ramming device is simply a big
wedge that flips things over... However you need both
speed and weight to generate the momentum to actually get
underneath anything else, and a lot of the lighter robots will
simply slide rather than be flipped. A block shaped ram can also be
effective for simply doing impact damage to lighter machine, but
again weight and engine power are required, and once a robot is in
a 10lb or heavier area remote control starts to become an essential
to avoid damaging people or hotel property...
The single most successful weapon so far has been string - tangling things up does tend to disable them. This is especially effective against the crowd pleasing robots with their mechanics in open view. Additional complications such as adding fish-hooks or similar refinements have also been tried, though without great success so far. The main difficulty is to avoid tangling up your own mechanism.
There are many other methods of attack; one I hope to try someday -
but would be happy to see anyone else use - is to bag/box
up other robots. A large cardboard box could be pushed
around with it's lid facing forwards and the flaps bent so as
to funnel things in but make if hard for them to get out again and
then you simply sweep things up. This could devastating with a bi
enough box and a remote control mechanism, but even without much
active control could still eat bimblebots, and be very popular
(especially with SMS...). I am currently working on a box carrying
chassis so may have something like this at 2Kon
A large tubular
bag - like those used to wrap Christmas tree - could also be used
to similar effect (with the add bonus that the struggling of poor
captured robots would be visible through the clear polythene), but
it is more difficult to mount and to capture with and would almost
definitely require remote control as the angle of attack would have
to be more carefully controlled.
Lifting/mechanically flipping other robots (re: Cassius from BBC Robot Wars II) is an elegant design idea, but it requires considerable power and a lot of weight (to either counterbalance the robot being lifted - or to give the momentum to actually get underneath the opposing robot). Using a ram or hammer type weapon is also possible, but sideways acting rams are recommended as a hammer/pick-axe is too likely to damage hotel dance-floors...
Rams though also need a lot of weight and power behind them or they will simply force two robots apart rather than actually doing any damage.
A number of other methods have sprung to
my (twisted?) mind... Firstly a small
fire-extinguisher that can be fired (possibly just by a
collision bumper) on short burst might be cold enough to crack
soldering on many electric vehicle - and would release an exciting
(but mostly harmless) cloud of CO2 - it might be possible to adapt
a broken sodastream to produce a similar effect,
but high pressure canisters are dangerous so take care people (and
always ask permission etc...).
A number of robots work well on wooden/tiled floor but have trouble with carpet, so a cloth laying robot might also cause some chaos (and possibly have artistic merit as well - tartan probably required for 2Kon...not sure what would be good for Blackpool in 2001...).
Flamethrowers are just too risky, and even cigarette lighter and related technologies are too likely to leave molten plastic on the hotel's nice wooden floor even if they don't cause a fire and so should be avoided.
Water pistols and similar are also likely to cause more mess than effect and should be at least discussed with the Cyberdrome crew before they are used in action. Party poppers could cause considerable excitement - loud noise with entanglement possibility - but they must be used carefully, and it would be well worth making sure they will not fire out of the arena.
Go build robots, bring them to the next Eastercon, please someone
else win the MBE (I can't face reading the book again...)...
Tim.
