Sumo Bots B/C [Trial]

starpug
Admin Emeritus
Admin Emeritus
Posts: 932
Joined: April 5th, 2008, 6:51 pm
Division: Grad
State: ME
Location: Waterville, Maine
Contact:

Re: Sumo Bots B/C [Trial]

Postby starpug » June 26th, 2010, 6:38 am

M-E-T-H-O-D MAN wrote:Sumo bots looks dope but we don't have it at our states.

It's gonna be an actual event next year.
Get your facts first, then you can distort them as you please. - Mark Twain

User avatar
ichaelm
Member
Member
Posts: 440
Joined: October 23rd, 2008, 3:10 pm
Division: Grad
State: PA
Location: CWRU, Cleveland, OH
Contact:

Re: Sumo Bots B/C [Trial]

Postby ichaelm » July 13th, 2010, 10:47 am

It seems to me that there are two main categories of design philosophies for the physical structure of the robot: heavy or light. Does anyone have any opinion / insight / experience about this?

AlphaTauri
Staff Emeritus
Staff Emeritus
Posts: 829
Joined: September 11th, 2009, 1:41 pm
Division: Grad
State: PA
Location: 04h 35m 55.239s, +16° 30′ 33.49″
Contact:

Re: Sumo Bots B/C [Trial]

Postby AlphaTauri » July 13th, 2010, 8:28 pm

My initial thoughts (having never done this event before) are that a heavy robot will make it harder for your opponent to push you out of the ring but might also limit speed. On the other hand, a light robot can move faster than a heavy one and you'll win in a draw under NYSO Trial rules, but also it's easier for your opponent to push you around. Hmm...what if you had a light robot, but with a brake of sorts to prevent your opponent from pushing you anywhere, like rubber stoppers attached to an arm that can fold out?

Or you could do this, but I think they'd DQ you for it.

Also, I have a question, what kinds of things do teams use to force their opponents out? Wedges, stuff to poke your opponent with, just ramming them with your sumo bot?
Hershey Science Olympiad 2009 - 2014
Volunteer for Michigan SO 2015 - 2018

]\/[ Go Blue!

User avatar
fmtiger124
Member
Member
Posts: 480
Joined: March 8th, 2009, 3:59 pm
Division: Grad
State: NY

Re: Sumo Bots B/C [Trial]

Postby fmtiger124 » July 14th, 2010, 6:17 am

I found that most common was some sort of wedge or surface that you could have your opponent end up driving up onto you and you then carry them out of the ring (a disadvantage of fast moving bots you can move too fast and end up on the other bot) or they would move in and try to pick the opponent up with some sort of really low surface. I did see a few that were just ram them out (like ours :| ) but I'm not really sure how they worked...our biggest problem if i remember correctly was it drove to fast and they kept driving it accidentally out of the ring also, lot's of wedges don't really have a side that's good for pushing....you'd have to somehow out maneuver the wedge and get behind it. As far as stuff to poke your opponent with I haven't seen anything like that and I don't think it would be too effective.
Image

User avatar
haven chuck
Exalted Member
Exalted Member
Posts: 343
Joined: April 28th, 2007, 2:35 pm
Division: C
State: PA
Location: Strath Haven

Re: Sumo Bots B/C [Trial]

Postby haven chuck » July 14th, 2010, 8:10 am

Does anyone happen to know how common ties are? As in, do they happen frequently (and thus it would be worth preparing to try to win the tiebreaker), or is the tiebreaker just in the rules for the rare occasion that there will be a tie (and so the tiebreaker isn't worth too much consideration over other things)?
2010 Can't Judge a Powder- NATIONAL CHAMPIONS
2010 Science Crimebusters- 3rd in the NATION

User avatar
jazzy009
Member
Member
Posts: 474
Joined: January 3rd, 2009, 1:12 pm
Division: Grad
State: -
Location: AJAX Chat

Re: Sumo Bots B/C [Trial]

Postby jazzy009 » July 14th, 2010, 9:22 am

ichaelm wrote:It seems to me that there are two main categories of design philosophies for the physical structure of the robot: heavy or light. Does anyone have any opinion / insight / experience about this?


Very deep, ich.
Yes those would be the main ideas when building this device under the trial rules. I hope they change a bit, though. For nats, the winning teams took the "heavy" approach (I believe, NY correct me if I'm wrong). However, there were others (such as my team) that went for the "light" approach. The guys on my team won the first few matches simply by running around the other opponents and hitting them occasionally so as not to be DQed. But, as I have already said, the winner (and top two, I think) were both the "heavy" approach.

Now let's take a look at the new "medium" approach. Lighter than heavy so you can run around the heavy guys, heavy enough to push around the lighter guys. All of that would involve a bit of luck though because you wouldn't be quite as fast as the light guys.
I don't even know why I'm explaining this...
Call me coach.

User avatar
ichaelm
Member
Member
Posts: 440
Joined: October 23rd, 2008, 3:10 pm
Division: Grad
State: PA
Location: CWRU, Cleveland, OH
Contact:

Re: Sumo Bots B/C [Trial]

Postby ichaelm » July 14th, 2010, 10:21 am

There are many tactical designs, but I can only think of 2 simple ones: wedge and ramming. Then there are many possible complicated tactical designs: lifting, hammering, etc. The majority or robots will probably be simple. And in a match between a wedge and a ramming robot, I am pretty sure a good wedge will almost always win. But this is assuming both are equal speeds. If one is significantly faster, I think the slow robot has no chance. By significantly I mean by a factor of 3 or something. So my idea of a good simple robot would be a fast wedge.

It is impossible for me to consider every possibility for complicated designs. But at least for now I'm not think about that yet.

In order to make a fast robot it needs to be light. But when you make a robot lighter, you lose traction proportionally. So I think the good robots are going to have wide wheels to compensate.

I think you only need to worry about the tiebreaker if your robot is slow. Fast robots will finish the match quickly.

This is all just my speculation btw. I have not built a sumo bot yet (completely).

Also jazzy's idea makes sense; maybe it's better to compromise between fast and heavy.

User avatar
lllazar
Member
Member
Posts: 839
Joined: November 19th, 2009, 7:20 pm
Division: C
State: IL
Location: Probably at my laptop, multitasking while on AJAX chat....

Re: Sumo Bots B/C [Trial]

Postby lllazar » July 14th, 2010, 12:47 pm

jazzy009 wrote:
ichaelm wrote:It seems to me that there are two main categories of design philosophies for the physical structure of the robot: heavy or light. Does anyone have any opinion / insight / experience about this?


Very deep, ich.
Yes those would be the main ideas when building this device under the trial rules. I hope they change a bit, though. For nats, the winning teams took the "heavy" approach (I believe, NY correct me if I'm wrong). However, there were others (such as my team) that went for the "light" approach. The guys on my team won the first few matches simply by running around the other opponents and hitting them occasionally so as not to be DQed. But, as I have already said, the winner (and top two, I think) were both the "heavy" approach.

Now let's take a look at the new "medium" approach. Lighter than heavy so you can run around the heavy guys, heavy enough to push around the lighter guys. All of that would involve a bit of luck though because you wouldn't be quite as fast as the light guys.
I don't even know why I'm explaining this...


I volunteered and was actually proctoring the Sumo Bot event at nats, mainly for experience since itll prob be an event this year, and i'll be doing it if it is.

The top two bots were NY bots. They were fast and relatively heavy (not anywhere near the 2 pound limit but a little more than a pound). From watching competition, ill say this much - if your gonna make a super light bot, give it up, you have no chance of beating a fast wedge. You also have to consider the fact that the top two were from NY, where this was an actual event and has been for the past 5 years or something....the kids doing the event were experienced and i have to say they were much better at driving than most of the other competitors. In this competition, if your looking for consistency, a simple, fast wedge that's relatively heavy (1 - 1.3 pounds) with an experienced driver can you get you far.
2011 Season Events~

Fossils (Regionals ~1st) (State ~6th)
Towers (Regionals ~1st) (State ~3rd)
Helicopter (Regionals -3rd gahhh) (State ~5th)
Wind Power (Regionals ~1st) (State ~3rd TIERED!)

Hooray for getting everything i wanted?

User avatar
ichaelm
Member
Member
Posts: 440
Joined: October 23rd, 2008, 3:10 pm
Division: Grad
State: PA
Location: CWRU, Cleveland, OH
Contact:

Re: Sumo Bots B/C [Trial]

Postby ichaelm » July 14th, 2010, 2:11 pm

Thanks for sharing your experience! I realize you can't just build for the minimum weight; you'll just get bumped out of the ring instantly, unless your opponent is really slow (don't count on it).

An another note, how do any of you plan on controlling motors? If not some kind of kit.

User avatar
jazzy009
Member
Member
Posts: 474
Joined: January 3rd, 2009, 1:12 pm
Division: Grad
State: -
Location: AJAX Chat

Re: Sumo Bots B/C [Trial]

Postby jazzy009 » July 14th, 2010, 4:36 pm

ichaelm wrote:Thanks for sharing your experience! I realize you can't just build for the minimum weight; you'll just get bumped out of the ring instantly, unless your opponent is really slow (don't count on it).

An another note, how do any of you plan on controlling motors? If not some kind of kit.


Definitely a kit, and as for pace, I believe 1st and 2nd were relatively slow. I remember a youtube video of NYSO Trial finals, the winner was a beast box.
In most finals under these rules, I imagine, you will see extremely large robots with extremely small and versatile ones.

...I can't find it, didn't look too hard, but I thought it was fairly easy to find.
Call me coach.

gh
Admin Emeritus
Admin Emeritus
Posts: 323
Joined: October 22nd, 2003, 5:47 pm
Division: Grad
State: NY
Location: CA
Contact:

Re: Sumo Bots B/C [Trial]

Postby gh » July 15th, 2010, 7:16 pm

I don't understand why you'd ever build light. There's a maximum mass for a reason. I designed under the assumption that everyone else would build to at least the pushing force given µ = 1.0 and given maximum mass, and at least about 10 ft/s of maximum speed (yes, per second).

The only reason you wouldn't be able to reach that kind of performance is money or technical capabilities. Beyond that point (getting a bot with good driving performance), it's all about practicing and maneuvers. No amount of fancy weapon mechanisms or these "light & fast" robots you fantasize about (seen too many cartoons, maybe?) can beat a fast, robust machine driven well by an operator with a lot of hours practicing.

You'd be wise to contact your closest regional team so as to build and practice your robots together. You will both benefit a lot from such a setup.
“Rather fail with honor than succeed by fraud.” —Sophocles

If you are looking to give help or get help:
Wiki > Forum post > AJAX IRC Chat or Java IRC Chat > PM > Email
Don't forget the Image Gallery, the OBB Archive, and the list of Exalted User medals.

User avatar
lllazar
Member
Member
Posts: 839
Joined: November 19th, 2009, 7:20 pm
Division: C
State: IL
Location: Probably at my laptop, multitasking while on AJAX chat....

Re: Sumo Bots B/C [Trial]

Postby lllazar » July 15th, 2010, 7:52 pm

gh wrote:I don't understand why you'd ever build light. There's a maximum mass for a reason. I designed under the assumption that everyone else would build to at least the pushing force given µ = 1.0 and given maximum mass, and at least about 10 ft/s of maximum speed (yes, per second).

The only reason you wouldn't be able to reach that kind of performance is money or technical capabilities. Beyond that point (getting a bot with good driving performance), it's all about practicing and maneuvers. No amount of fancy weapon mechanisms or these "light & fast" robots you fantasize about (seen too many cartoons, maybe?) can beat a fast, robust machine driven well by an operator with a lot of hours practicing.

You'd be wise to contact your closest regional team so as to build and practice your robots together. You will both benefit a lot from such a setup.


This event is 60% bot, 40% driver...your gonna need someone with some mad joystick skills, just putting two kids with robotics knowledge won't get you anywhere...i noticed something at nats, this event can trip people up because it's so unlike all the other events. This event requires skills other than the ability to study harder than everyone else - heres an example of what i saw,

A lot of drivers weren't concentrated. I saw that they struggled to think on their feet, and if you can't figure out what to do quickly, a decent bot with an experienced driver will take advantage of an and every opportunity present to them.
2011 Season Events~

Fossils (Regionals ~1st) (State ~6th)
Towers (Regionals ~1st) (State ~3rd)
Helicopter (Regionals -3rd gahhh) (State ~5th)
Wind Power (Regionals ~1st) (State ~3rd TIERED!)

Hooray for getting everything i wanted?

User avatar
ichaelm
Member
Member
Posts: 440
Joined: October 23rd, 2008, 3:10 pm
Division: Grad
State: PA
Location: CWRU, Cleveland, OH
Contact:

Re: Sumo Bots B/C [Trial]

Postby ichaelm » July 15th, 2010, 8:25 pm

gh wrote:I don't understand why you'd ever build light. There's a maximum mass for a reason. I designed under the assumption that everyone else would build to at least the pushing force given µ = 1.0 and given maximum mass, and at least about 10 ft/s of maximum speed (yes, per second).

The only reason you wouldn't be able to reach that kind of performance is money or technical capabilities. Beyond that point (getting a bot with good driving performance), it's all about practicing and maneuvers. No amount of fancy weapon mechanisms or these "light & fast" robots you fantasize about (seen too many cartoons, maybe?) can beat a fast, robust machine driven well by an operator with a lot of hours practicing.

You'd be wise to contact your closest regional team so as to build and practice your robots together. You will both benefit a lot from such a setup.

I tried to do some calculations to try to see if you could get a ~2 kg sumobot to accelerate from 0 to 3 meters per second within a ~3 meter distance, running with reasonable voltages and internal resistances. But I realized it completely depends on the motor and gearing system, and I don't have the time to spend figuring all the math out (I'm not that great at math). But I've already done some tests on motors that I have, and they just don't come close in terms of the combination of torque and maximum RPMs you would need. Of course, most of my motors come from printers or RC cars and stuff, so it's probably a terrible comparison. One idea I came up with was some sort of gearshift, to keep the motor running at its most effective torque and changing the gear ratios to compensate. But I really don't think that's what you had in mind. It's such a short distance that a gearshift would be practically useless (correct me if I'm wrong). If you have motors with that kind of combination of torque and maximum speed, please inform me where I can get some!! :P

Because it is definitely true that a 2 kg robot going 10 feet per second will dominate over a 1 kg robot of similar speed and design. The problem, as I see it, is matching high speed with maximum mass. I have not finished my first sumo bot yet, and you've built one or more already, so I figure you've figured this out already. Is it really just about getting the most powerful motor money can buy? (and matching it with an effective wedge or something). I hope not; up until now I've been building my robot completely out of salvaged parts from broken stuff we have lying around and basic electronic components (transistors, etc.). Except for that radio set, which I've temporarily borrowed from a really cool old handmade tug boat we have for no reason (we live nowhere near a lake or anything), and I promised to put it back in the boat as soon as I get a legal set. If I need to get 100$ motors or something it'll ruin my cheapness :roll: . BTW this is just for my first design over the summer. For the real one we're definitely going to use more specialized parts and digital electronics.

Sorry, back to my main point: Can you elaborate on what kinds of motors you used, or why you do/don't need super expensive motors?

PS I understand the value in practicing :D Especially with something as fast as this is supposed to be.

gh
Admin Emeritus
Admin Emeritus
Posts: 323
Joined: October 22nd, 2003, 5:47 pm
Division: Grad
State: NY
Location: CA
Contact:

Re: Sumo Bots B/C [Trial]

Postby gh » July 16th, 2010, 5:35 pm

I had 550-size motors, where the "-size" is a quasi-standard for a motor whose can is about 55mm long. They are easy to find, but finding a good gearhead or gearbox for them will be hard. That's why I think power tools are a much better source of powerful motors with gearboxes. Cordless drills will even have decent batteries to rip out. My initial blueprints for my senior year bot involved DeWalt cordless drill motors (roughly 1.2kg for a pair; you can imagine their size) and half of a DeWalt 36V LiFePO_4 pack, but that was too insane even for me.

What I used (see earlier in the thread) was far overkill even for my somewhat lofty goals. Remember that with powerful motors you need powerful motor controllers and batteries. My batteries could give about 1kW (2kW peak), though my first motor controller spazzed out at .25kW or so because of overcurrent protection. An accelerating bot consumes a metric heapin' of power, I'm afraid.

Also, for my current project in college, my buddy and I are making my own sensored brushless motors because the torque and peak power we need far exceed what our budget can provide.
“Rather fail with honor than succeed by fraud.” —Sophocles

If you are looking to give help or get help:
Wiki > Forum post > AJAX IRC Chat or Java IRC Chat > PM > Email
Don't forget the Image Gallery, the OBB Archive, and the list of Exalted User medals.

User avatar
ichaelm
Member
Member
Posts: 440
Joined: October 23rd, 2008, 3:10 pm
Division: Grad
State: PA
Location: CWRU, Cleveland, OH
Contact:

Re: Sumo Bots B/C [Trial]

Postby ichaelm » July 17th, 2010, 9:46 pm

A while ago I did some research on different kinds of batteries. If I get any motors more powerful than the ones I have access to now, I'm going to need less internal resistance. An average NiMH or NiCD cell has an internal resistance of around 0.06 ohms. This means that its maximum usable power is about 6 watts per cell, or 72 watts at 14.4 volts (12 cells). And I'm pretty sure those kinds of batteries aren't meant to carry over 5 amps. I guess the logical next step would be LiPO, which you talked about. SLA is just too heavy :lol: . I'm not sure if I'm ever going to need that kind of power though. Like you said, at some point you're limited by traction and other factors. With my motors. it would be fruitless to go over 50 watts. Although my motor controller is custom built (actually it's still in the breadboard) so if I really need to some day I can swap in some even bigger transistors. I think the only way you could actually benefit from over like 200 watts of power would be with sticky wheels, or treads (um no). And I hope they outlaw adhering to the floor; it can just get out of hand.

I was thinking: what do you guys think about brakes? Has anyone tried a physical braking system? Pretty sure you'd only ever need physical brakes if your motors were really weak (or not geared down enough). What about shorting the motor terminals? Is it effective/safe/necessary? Just trying to get some input from more people with more experience.

EDIT: see next post...
Last edited by ichaelm on July 24th, 2010, 10:37 am, edited 1 time in total.


Return to “2010 Build Events”

Who is online

Users browsing this forum: No registered users and 1 guest