Let's Make Robots!

RoboGames Sumo

I'm still very new to robotics, and there's no way I could just build a sumo robot right now.

But when I said I wanted to run a marathon a couple years ago, I wasn't a runner, and I just worked my way there slowly but surely, and finished the race.  Moral of the story; I'm much more motivated if I have goals.

So anyway, I want to enter the sumo robot competition at RoboGames in 2009.  I'll be building simpler robots and learning more about robots in the mean time, but I figure it's never too soon to at least be thinking about it.  Also, having that as a goal will kind of give me a reason to keep learning and improving.

So I'm wondering if anyone here has ever competed in a sumo event.  If so, can you share your experience with it?  Do you have any advice for a first time participant?

Actually, if you've ever entered any robot competition, I'd love to hear about your experience with it, just out of curiosity. 

Also, there are several weight classes and I'm wondering which one is the "standard" or the one that is most popular.  Here's a list of the sumo weight classes, plus all the other events.  Sumo is on the lower left.

http://robogames.net/events.php

Lastly, I'm going to RoboGames 2009 whether I enter a competition or not, because I live in California it looks like a lot of fun.  Will anyone else be attending?

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

There are a number of different sumo events. I've had 3 minisumo robots, which appears in Robogames as 500 g Autonomous. I've watched a few of the 3kg autonomous matchs. THe 3 kg events were started in Japan, and coulod see some very tough competitors come from ther. Their events copnsist of school teams competing, going through many teams(500? 1000?) to have their top 5 sumos come over to compete in autonomous and RC 3 kg events. The robots are limited to 3kg weight and 20cm by 20cm in size, and in the past had made use of vacuum, to hold them on the 5' (154 cm) ring better. Used to have a video of a sumo operating a search pattern on a vertical wall. I think current rules are geared towards magnetics, as the ring material has changed to steel I think. Sumo speeds can get to 5 meters per second The US primarily started with the 500 g sumo from efforts by Bill Harrison a number of years back. Minisumos are 10cm by 10cm and compete in a 30" ring. (77cm) no vacuum allowed, autonomous only. The 1 kg is a lego addition since the bots built with the mindstorms stuff tended to be a little heavy or large for minisumo, but too light for full sumo. The 100 g is microsumo, I think 5 by 5 cm, and then 25 g nanosumo, 2.5 cm by 2.5 cm. The last 2 are getting just way to small, but some have funwith them.

Minisumo is much easier cost wise to get into, and is a lot of fun. Most competitions were double elimination (the robot had to lose twice to be knocked out). The robot must be programmed to wait 5 seconds before moving. Then it's simply use the best pushing or strategy to get the other robot out of the ring. Simple things like keeping your robot in the ring can help the robot win the match. I've seen minisumos that blindly drove off past the other robot. Next is trying to find your opponent. And then maybe track your opponent and decide what to do next. Things like weight, traction, speed, and precision are good to keep in mind in building and programming. Here is a couple pics of my first minisumo, that actually did ok in competition a couple years, til I went against the originators in Seattle. Sarge was a fun little robot, that I still have, though he needs work at the moment.

P4300003.jpg

P4300005.jpg

Sarges' rubber washer wheels got replaced by silicon sealer at one point, by slowly squeezing it out of the tube while the wheels were rotating, letting a layer dry, then repeating. Lead weight was added to get as close to 500 g as possible. The next minisumo was Nestor, based on a Mark II chassis, and most recently Grendel, scratch built on a Tamiya twin gearbox. There are a number of videos online of minisumo and sumo matches, that would be very educational.

Other competitions I've been in have been local events with the club, which don't appear to be similar to anything at Robogames. These were  Quick Trip, T-Time, Line Following, somewhat simple contests. I've recently been interested in what has been called Robo-magellan, an outdoor robot navigation event that I have not competed in yet. There are a number of these competitions throughout the US. They appear to have that at Robogames too. Firefightling is another that might be interesting.  Even balancer race too.

That's great, Robologist!  A lot of good info ... I really really appreciate you taking the time to give me such a thorough answer.

It sounds like 500g is the way to go, then.  Yes, 3 kg could be very expensive, and it sounds like it might be more competitive too.  I'm sure I won't win any tournaments, but winning a few matches would be nice.  :)

What's this about squeezing glue onto the tires??  I'm amazed that they allow that, and that you managed to actually get it working!

Since I posted the original question I've been doing some research online (http://www.robotroom.com/SumoRules.html was VERY informative), and I'm already getting some ideas for a design.  I read that you can stealth the bot from IR and ultrasound by using black surfaces or even soft materials to absorb ultrasound.  I also read that your bot can expand once the match starts.  So here's what I was thinking: have a bot that is as invisible as possible to sensors and when the match starts, have a reflective object (still attached) come out on one side.  Then, try to get the other bot to take the bait and push them off the edge from behind, where they are less likely to have a pusher thingy that's close to the ground.

Have you seen anyone try this before?  If not, do you think it would work, or would it be more trouble than it's worth?  Perhaps my first sumo bot should just be a simple "I'ma gonna ram you" design, and save this idea for later.

Before you had multiple bots, how did you test out your sumo to see if it would be any good?  Do you have it push bricks off tables or something?  I unfortunately have no robot clubs anywhere near where I live, so there's no one to practice against.

I've seen a couple of robots that did "distractions" waving a flag to the side, as they used stealthy coverings and such. They did ok, but generally the best stealth was making a low robot, that moved quickly.

FOund a general list of minisumos, pretty cool : http://www.er-online.co.uk/minisumo/wiki/index.php?page=Gallery

Here are a few top competitors I remember hearing or seeing : 

Goliath and Solo 

Nemesis II  

Seeker II 

Solo and Enders Wraith

These have a pretty common theme of low, fast, and custom made tires of low durometer urethane. 

This isn't to say that other minis don't win, Dave Hylands Marauder did very well for a long time and it is simply a highly tuned and well programmed Mark III kit robot.  It seems the Solarbotics Sumovore kits do very well too. With either of these, it would be good to add the silicon or urethane tires.

On Sarge,  I used silicon sealer, that dried but was still soft and gripped the ring better. When building, I had to run the wheels and apply a layer, let it dry a few minutes, then apply another layer. Took about 30 minutes, but wound up with tires similar to the urethance grip of some others. Sarge was a very slow robot, but persistant. 

And yes, just ordinary objects to push, soup can, brick, bowls. I remember hearing Goliath tested out pushing a meachanics toolbox around.

 

Great links, Robologist.  I've been reading robot sumo pages for the past hour.  :P

The page on Goliath and Solo was especially helpful.  I think I'll build something like that.

I have a friend here in town who does metalworking ... maybe he'll agree to build the shell .....