To get started on Robot Wars UK you will need to contact Mentorn the TV company and join the Robot Wars Club, it will cost about £13. Information on this can be found quite easily on their web site at http://www.robotwars.co.uk, make sure you indicate that you want to build a robot. If you are aged under eighteen you will have to have an adult on the team.
Other Web Sites
Explore the Net, there are a lot of Roboteers sites which have helpful information in the form of Hints and Tips; a good starting point would be the Dangerous Machines Web Site at http://www.tinweb.com/robots which has quite a lot of technical tips and the most comprehensive list of links to other Robot Wars sites and manufacturers.
Exotic designs and walkers
If this is your first robot, don't build a walker, they are quite complex either with a mechanical linkage between the legs, or in their software. Walkers are not competitive with the rollers. The failure rate of teams building walkers is very high. For Jim Struts we built a very crude prototype out of Lego to prove that the walking would work. The mechanics behind the legs had no resemblance to Jim Struts what so ever, but it did demonstrate that the method of walking would work.
With Jim Struts we took some big risks, we knew we could build him and make him walk but we did not know if we could make him within the weight limit or get him to walk in time for the filming. We did not make the weight limit and could of done with a lot more time working on the software. We made lots of mistakes, which I hope are explained in our web pages.
Tank Tracks and large number of identical items
Tank Tracks, I, personally, would avoid them. You need very powerful motors to drive the tracks. The tracks have to slide on the ground sideways to be able to steer so you can not have as good a traction as rubber wheels. There is a second issue with tracks, they are made up with many identical parts which when you sit down to construct can be very time consuming. If it takes an hour or more to make one tread or other item, and you need 100 then it'll take over one hundred hours! Don't bother unless you can get some ready made.
Design your Robot, draw a sketch (it need not be too detailed), an outline for space planning and sizing is all that is required. Try not to get bogged down in too much detail. Part of your design process should include an analysis of the weight. This can be easily done using a spreadsheet program. Create a list of all the key parts of your robot, and estimate or use the known weight for each item. As you find the actual weights start replacing the estimated values, this will help identify at the earliest opportunity when your robot starts going overweight. This will hopefully allow you to take corrective action in time. Included a line for sundries to represent odds and ends such as nuts and bolts.
I have observed that some roboteers use a standard 40MHz radio equipment and use mechanical linkages from the servos to variable resistors for the speed controllers or operate micro switches mounted close to the horns of the servos. This is a nice quick way of making the radio equipment drive the robot but not as effective as an electronic interface between the radio equipment and the speed controllers. It is worth exploring other Robot Warrior's web site's as some do offer these interfaces for sale.
This bit is not easy, and not entirely necessary, but is an interesting exercise. Try to work out how long it will take to build identifiable parts of your robot e.g. gear box, chassis, and other processes such as designing and testing. Then place these in an order in which they need to occur to build the robot. From this you should be able to work out a crude timetable of when things should happen. Hopefully from this you can start to predict if you are falling behind in your construction. Remember that if you have more than one member on your team some of the tasks can happen at the same time.
Despite having a plan/timetable things will not go 100% as expected, so you will probably have to revisit you plan to take this into account. As soon as you detect that it will not happen as you expected, reflect this in your plan, and hopefully you will see a way of rescheduling tasks to meet the project deadline or predict that it can not be achieved.
The above description is very crude, project planning can nearly be an exact science and I am sure there are many books on it, but bear in mind that it works best on medium to large projects such as building Motorways, ships, moving offices, putting a man on the moon, making a Robot Wars TV program/contest, anything which the tasks can be measured in terms of days and weeks. Some projects can be planned in terms of minutes. This would apply to something like replacing a part to a complex machine which runs 24 hous a day, and cost millions to have shut down. It allows you to mimimise the shut down time.
This is TV
It is worth remembering that Mentorn are a TV company and are running the competition to make a TV program. They will be looking for robots that are effective, have interesting weapons, and look good. If you have watched every program so far, you may have noticed that some of the robots begin to look same, I call this the aluminium box on wheels effect. The production team have seen a lot more robots than we have so this effect will be more pronounced for them. Make your robot interesting not an aluminium box on wheels.
The House Robots
The House Robots should be treated as hostile mobile scenery. Whatever you build Mentorn will have a robot that can defeat it, so don't try. Your aim is to beat the other competitors.
Time And Money
These really do equate if you have loads of time then you can spend time looking for second hand items and modify them for your purposes. If you have little time you will probably need to spend money buying ready made or fit for purpose items.
So if you want to make a robot for next to no money allow plenty of time to find materials or re-working existing materials.