Society of Robots - Robot Forum

General Misc => Misc => Topic started by: Kohanbash on October 16, 2007, 09:59:22 PM

Title: Robot Characterization
Post by: Kohanbash on October 16, 2007, 09:59:22 PM
Hi
I am starting this thread in response to http://www.societyofrobots.com/robotforum/index.php?topic=2159.0
As far as I know there are no threads which talk about how to describe and characterize a robot.
Title: Re: Robot Characterization
Post by: Kohanbash on October 16, 2007, 10:03:56 PM
Just to start this off here are some common ways to characterize a robots performance:
1. Mass
2. Size (Dimensions)
3. Center of Gravity
4. Torque
5. Maximum ascent and descent angles
6. Tip-over angles
7. Towing capability
8. Power Consumption (idle and in-action)
Title: Re: Robot Characterization
Post by: Ro-Bot-X on October 16, 2007, 11:54:20 PM
Just to start this off here are some common ways to characterize a robots performance:
1. Mass
2. Size (Dimensions)
3. Center of Gravity
4. Torque
5. Maximum ascent and descent angles
6. Tip-over angles
7. Towing capability
8. Power Consumption (idle and in-action)

I am asuming we are talking about sumo robots, so:
1. Make it the maximum allowed. Use a digital scale. Use what ever as weight to get it just right to be admited.
2. A wide robot will turn slow, so make your robot longer, not wider is possible.
3. Keep COG low, something like 2 thirds over the wheels axle and one third on the plowing wedge.
4. You have to figure a torque-speed-friction formula that will keep your wheels spinning when going against a wall, but not verry fast spinning. Someting like one rotation in 2-3 seconds. Use wide wheels, rubber or silicon, something a little softer than hard rubber. If you use neoprene, coat it with silicone.
5. You don't need that in a sumo robot.
6. Same like above.
7. In sumo it's pushing force. That will be figured out in step 4.
8. Power... never calculated it.

That will be from my sumo contests experience, all with Lego robots, some contests using a pure mechanical robot without a microcontroller and some contests going against regular robots (mini sumo). Back with pictures soon.
Title: Re: Robot Characterization
Post by: Admin on October 17, 2007, 09:13:42 AM
A few things left out . . .

velocity/acceleration

And the much harder to measure factor:
strategy (and the design required for that strategy)

Some people use simulation . . . I just made a 2nd RC robot to have bashed around by my sumo bot - the faster it got bashed, the better my sumo design was ;D
Title: Re: Robot Characterization
Post by: Kohanbash on October 17, 2007, 09:58:01 AM
Ro-Bot-X

I was referring to any type, not just sumo.
For example today I am going to be testing a robots draw bar pull.
Title: Re: Robot Characterization
Post by: Ro-Bot-X on October 17, 2007, 11:32:11 AM
Without strategy, the sumo robot is like a mobile brick.
Active sensors make the difference. Active as in scanning for the enemy. Admin's robot Stampy is an excelent sample of sumo robot. I am not sure I can beat it with my Lego robots using just passive sensing.

Now here are some pictures of a few Lego Sumo Robots I have built over the years:

http://www.flickr.com/photos/11436019@N06/sets/72157602474815967/ (http://www.flickr.com/photos/11436019@N06/sets/72157602474815967/)

(http://farm3.static.flickr.com/2083/1602262346_189d2557cc.jpg?v=0)

(http://farm3.static.flickr.com/2256/1602263184_238c8d5a77.jpg?v=0)

(http://farm3.static.flickr.com/2315/1602257810_8157b9e916.jpg?v=0)
Title: Re: Robot Characterization
Post by: gamefreak on October 17, 2007, 06:05:34 PM
LOL, got enough traction on that sumo? 24 wheels, lol, beastly
Title: Re: Robot Characterization
Post by: JesseWelling on October 17, 2007, 09:38:26 PM
It's going to have a lot of traction when turning in place too :-\
It would probably turn faster if you made the wheels not so close to the center line.
OTOH even if it gets blindsided it's probably not going far  ;)
Title: Re: Robot Characterization
Post by: airman00 on October 18, 2007, 05:17:32 PM
Can sumo robots have spiked wheels , so that the spikes actually stick into the table?? Is that against the rules?
(http://img.nextag.com/image/Swisher-Drum-Style-Spike/1/000/005/336/634/533663408.jpg)


these questions are more techinical and legal questions than electronics related questions.

If you do spikes be careful to pick it up!    :P
Title: Re: Robot Characterization
Post by: Admin on October 18, 2007, 06:32:56 PM
Quote
Is that against the rules?
yes. sticky wheels are also against the rules.
(cant damage or modify the playing field in any way)
Title: Re: Robot Characterization
Post by: Ro-Bot-X on October 18, 2007, 11:14:45 PM
It's going to have a lot of traction when turning in place too :-\
It would probably turn faster if you made the wheels not so close to the center line.
OTOH even if it gets blindsided it's probably not going far  ;)

Hehe, the robot had 2 Lego high speed RC motors, one on each side, powered from a separate 9V AA battery box. The Lego RCX controller was using regular geared motors connected to Lego electrical switches to control the RC motors going forward or backward. When turning, the robot would shake a lot sideways. Lifted from a side, the robot would spin aroud with the pusher, wearing the wheels on the oposite side. I actualy got pieces of rubber of the ring after a match. Oh, the weight was 2 pounds. If the robot would catch it's enemy front sided, it would push it out with no problem at all!