Formula MiniSumo

[slurp]

Hi,

I do like the development of minisumo but I can't help thinking that some new competitors are discouraged by the thought that they can't compete. Some events run multiple competitions with novice categories to give that opening but I've been considering a formula with a view to an easy / budget competition, in much the same was as Formula Ford sets level in motor racing.

I'm hoping that a formula will support technology in schools as well as breaking down the daunting part of the competition where experienced builder may be spending a small fortune on robot. For our international readers, I apologise that I've biased these component to the UK Educational and Hobbyist Market with the base specification. I trust that this opens a debate....

Control:-
PICAXE-18X micro controller - £4.35 (inc. VAT)

Drive:-
Economy Gear Motor - 2 x £2.62 (inc. VAT)

Recommended Support Circuitry:-
PICaxe 18 Pin High Power Project Board - £6.93 (inc. VAT)
L293D motor driver - £3.53 (inc. VAT)

This leads us towards this sort of configuration (L293D and PICaxe sockets shown empty) :-



In terms of size, this is approximately 90mm wide with 12mm spacing between gear motors, 75mm long and these economy wheels are 38mm diameter. I guess the gear motor is the most levelling part of the specification, everything else makes the entry easier.

A PIC16F88-I/P is the base controller behind the PIXaxe 18X but the PICaxe give an easy entry level to controllers. The gear motor is a low cost element manufactured by Vigor Precision but is also know by the Solar GM motor series. Where people are gaining confidence I expect that they'll start to consider other controllers and other drive systems but will then be moving out into the competition as a whole with creeping expense and performance.

I've started a wiki page to document the formula (http://www.minisumo.org.uk/wiki/index.php?page=FormulaMiniSumo), and a demonstration build, but I'm open to debate and change. I'd like to make this low cost and accessible but also reasonable competitive.

I'd welcome you comments.

Best regards,
Colin

Wiki Link Edited to fit new format

[sap1975]

I’m still a big fan of the Formula idea.
Although i would have looked a taking it in another direction

I assume you want to clarify a class between Lego and the “high end” stuff. That sound to me like a great idea but the way i see it the problem would be to have the established competition events adopt it. Have you had any thoughts on how you would do that?

Last but not least it might be an idea to make more specific rules on what is and is not allowed such as disallowing LI-PO’s, defining a maximum motor size, maximum voltage and number or type of sensors instead of defining the actual parts. It would leave room for a little bit of ingenuity

Anyways… just thoughts.

[slurp]

While we see Lego as a class it's not really a cheap entry level, yes I'd like to make this some at the low end - hopefully giving all the kits a good run for their money!

In terms of establishing the sumo, I'm working this up with a view to using it with local schools. Ok, I'm running late again... one or two with proposed specification will appear at the National Space Centre event in mid May.

I've been considering sensor or budget limits but thought voltage a useful variable. Some of the starting problem is where to start, what to use... a few specifics and examples can help pass the blocks - I guess that's why the kits sell well, even if they can be expensive. The biggest leveller has got to be the gear motor but I'm not sure it's the only one that should be used. I think we'll always find room for ingenuity.

regards,
Colin

[slurp]

Thinking of a conversation with Damian (Techbotics), we where talking about another event but looked at a points system. This could be devised to limit the overall performance but with flexibility to improve one aspect of performance while limiting another.

This could get very complicated when you start considering how you score various key components, or have to judge components against a spec. I guess this comes back to limited component / specification sets.

still thinking....

[RiO]

I propose to make this class based on popular stock kits (Stock Class).

It will allow for competition:
- bots from kits which are generally available for under [$100-150] with at least 50 kits sold per year
- UNMODDED hardware
(perhaps some minor tweaks will be allowed, but not to change motors, voltage, and sensor system)
- custom software allowed.

Benefits:
- relatively easy entry
- avoid hassles of hardware development, focus on software
- incentive for manufactures to make cheap and efficient kits

The list of qualified kits may initially include:
- Solarbotics Sumovore
- MArk III
- [I know less about other current kits]
Additional kits will be included by the tournament organizers.

[slurp]

I propose to make this class based on popular stock kits (Stock Class)

I think this would be a seperate class, I don't want to kill that discussion but think that it should be a seperate thread. The budget and scope of build doesn't really corespond to the formula proposed, particularly to inspire thought with regard to build.

There are a few events that hold multiple competition, a beginners and master class. Such that competition winners and sometime non-stock builders can not compete against the new starters but the top performers in the beginners are free to enter the master class.

I think I should move this discussion to discussion on tournament rules (starting a second for Stock Class)...

There's a parallel discussion on the Robot Club Forum (Hawaii) relating to High School Mini-sumo workshop - best kit? which includes elements of interest. Some of this discussion relates to budget, how you account for it and using it as one of the limiting factors.

regards,
Colin

[RiO]

Slurp,

I read the Hawaii forum discussion and have the ultimate cost reduction idea: virtual mini-sumo.
There will be some simulator software, which can imitate performance of a standard mini-sumo robot (or a selection of common kits).
It will take the code and placement instructions from the user and then emulate the processor actions, calculate robot movement and feed the sensor data to the virtual processor. The result will be micely renderedin 3d.
Typical current desktop computer will easily handle this.

I have seen such simulators for robot football (MiRoSot). that task is way more complex.

This way a school kit for a full class will include a group software license (hopefully for $0), and 4 standard live robots. The kids would test their programs virtually and only use the live robots only to cofirm their algorithm performance and to take them for outside of classroom events.

[slurp]

The trouble is that virtual minisumo just isn't real and it doesn't give the same hands on experience.

How many random effects appear in the simulator? Can we account for batteries, dust on tires, lighting? Ok, it also takes a good teacher to recognise and communicate these effects in practice too. Building, and building from a non-kit idea, can teach a lot - especially where the failures are considered.

I would expect that most $50 builders would relish the opportunity to test there metal against a $100 stock kit, or even a $400 build for that matter. I expect competitions will always cater for all but the Formula idea is about getting that first taste for competition on a level(ish) basis, I really want these builders to move on and take on the $400 build and win... maybe they'll spend that in the end but they'll be spending on what they learned from a $50 budget.

best regards,
colin

[Techbotics]

Hi Colin et all

I also believe this is a great idea to get schools etc involved in MiniSumo.. However can we please keep the controller choice open.. I am willing to say limit it to 8 or 16 bit and 16mhz (for example) and also to impose a memory limit if need be, but personally I am a C guy very happy with the ATMEL controller and my set up in Eclipse and would really rather not have to learn to use another type if it can be avoided.

Happy with stock drive train, sensors, tyres etc though..

Opinions..? Also Colin do you have the latest recipe for the mini sumo formula available for perusal.. I see the above but have you considered sensors yet at all..?

Cheers

Damian


http://www.techbotics.com

[slurp]

As posted earlier in the thread - http://www.minisumo.org.uk/wiki/index.p ... laMiniSumo

I've not changed this yet but had some interesting discussion with the MicroMouse chaps at the NSC event, particularly on the goal of the formula (a flexible but more uniform low cost beginners competition) and practicality of cost limitations. Does the cost limitation assist the goal or does it bring an accounting nightmare?

How would you account for expensive machine parts donated by a sponsor vs. similar that might have been produced in a school workshop, or even similar parts that are lower quality where the tooling isn't available to the builder. Is it better to consider limiting only key components that level the competition?

I believe that the controller should be open, the recommendations are simply that. Main aim was to guide new users to low build cost and simplicity of construction or programming (eg. BASIC or FLOWCHART). There are a couple of object sensors I've considered but again as recommendations.

I can't help but think that some sort of accounting for cost is required, or some sort of points basis as we have previously discussed off forum. I still think the stock low cost motor/gearboxes are key, possibly with a voltage limitation (recommend minimum 6V but say 9V max).

best regards,
colin

[Techbotics]

Hi Colin

I'm not sure cost is the way to go for the "Formula" MiniSumo.. as you have already pointed out it leaves many loopholes the rules would need to take account of such as donated parts, 2ns hand parts, home made (what is home made anyway.. is it still home made if I have a CNC mill in my garage (I wish... )) etc etc.

If we follow the Formula Ford model we could either :

(a) restrict the bots using general rules (for example: The processor much be 8bit CPU, max 4K memory, max 16Mhz, battery will be standard chemistry and a max of 9V and ?? mAh)

Or

(b) Specify approved parts, ie The processor must be one of: PiCAXE, ATMega8 etc.. Battery must be a standard 9V PP3 (any alkaline brand). (the approval process could become tiresome with this method.)

While all the time remembering the whole point of the formula is to keep costs down, this should govern the rules or parts selected above.

Regarding donation of exotic material or home machined parts, I am not sure these have a place in "Formula" MiniSumo and instead are more suited to the “Unlimited” class.

Open to opinions.. what do you think.. I am happy to help drawing up the rules if you want, it will be great to see some "Formula" bots at the upcoming MiniSumo events.

BTW I have just received my budget GHM's and some budget wheels for my first 2 Formula bots.. (one big wheel one small wheel).

Did you machine a slot in the wheels to make them fit the GHM or just a hole and some glue..? Or did you attach them some other way..?

Blimey this gives me the 2 SAMrI complete and 3 other bots on the go.. It looks like I wont be doing any biped or BEAM work for a while.

Cheers

Damian

http://www.techbotics.com

[slurp]

If we follow the Formula Ford model we could either :

(a) restrict the bots using general rules <<SNIP>>

Or

(b) Specify approved parts <<SNIP>>

While all the time remembering the whole point of the formula is to keep costs down, this should govern the rules or parts selected above.

That's it but to avoid killing ingenuity I don't want to be too restrictive. I guess, like Formula 1, the rules can be reviewed when something appears that un-hinges the balance.

A range of gear motors seems sensible (as suggested) but is this enough in itself?

Regarding donation of exotic material or home machined parts, I am not sure these have a place in "Formula" MiniSumo and instead are more suited to the “Unlimited” class.

Again this is in the broad spirit - are we suggesting hand tool construction? Though I would suggest that Pillar Drill should be allowed - common in schools and can be found as a budget(?) tool (£25 - ref: http://www.minisumo.org.uk/forum/viewto ... ?f=1&t=231 )

Open to opinions.. what do you think.. I am happy to help drawing up the rules if you want

Isn't that what we've started here

Did you machine a slot in the wheels to make them fit the GHM or just a hole and some glue..? Or did you attach them some other way..?

Ah, that's noted here http://www.minisumo.org.uk/forum/viewtopic.php?f=5&t=76 (some time ago... that why I need to push stuff out into the wiki!) I'm going to add some plates to the inside of the wheel to pick up on the shaft flat for better torque transfer - will work easier on the small wheels compared to the big (smaller one I've turned to use the outer boss too).


Probably off topic, I'm looking at getting some (simple) boards made for basic control, drive and regulation to assist a school club - 3rd party design (so a good job - I hope!). Moulds for some wheels to fit these low cost Gear Motors is a possibility.

regards,
Colin

[slurp]

Missed a bit - I don't want to get to bogged down in battery type, I think this should be open for rechargeable and not necessarily just Ni-MH or Ni-Cd as Li-Poly may suit too. However if we stick to a Max Voltage and a specified Motor Driver such as the L293 - maybe not the most fantastic device but it's cheap and common! (is this looking like a Sumovore fits?!?)

regards,
Colin

[Techbotics]

Hi Colin

Not sure about Li-Poly due to cost.. but perhaps you can get a better price on them than I can.. lol..

I would be interested in looking at the control board spec as I am in the process of having some made too, perhaps we can get some economies of scale here.

Cheers

Damian

[slurp]

Not sure about Li-Poly due to cost.. but perhaps you can get a better price on them than I can.. lol..

Hi-Model shipped mine to be at about £5.70 per battery and £6.45 for the charger - shipping £13 (Ref: http://www.minisumo.org.uk/forum/viewto ... ?f=11&t=35 ).

I think the average AA battery chager (with four AA's) is about £25 seems similar to me... buying Li-Poly in the UK is often twice this

regards,
colin