DiscoBots Omni Drive over Bump

[lynca]

DiscoBots testing their omni drive robot on Wednesday, it worked well over the bump. These are 8" AndyMark Omni wheels with standard toughbox configuration.

http://lamar.discobots.org/node/8

More wheels will be added to this drivetrain in the future, but this was a good proof of concept.
Suggestions, Comments ? Let us know what you think.

[racer6]

looks good

[PerpetualMotion]

Why omni directional wheels if you can not go in omni directions

[nickd2012]

Why omni directional wheels if you can not go in omni directions

"More wheels will be added to this drivetrain in the future"

[LostProphet]

i think Omni wheels would work if your going with a front wheels drive.

[Geek 2.0]

Where do you intend to put the other wheels?

[PerpetualMotion]

Unless you have a 4 wheel drive robot with two traction wheels and two omni wheels, I think this is a pointless experiment. Whats the point of having 4 omni wheels if you don't move in omni directions. If you are going to add wheels, then its sounds like you're going to have 6WD, which has a much different dynamic for going over bumps then 4WD, so nothing from this experiment can be translated. If you are going to do some sort of holonomic drive, then you are going to be losing traction and again this experiment becomes useless.

[Geek 2.0]

Unless you have a 4 wheel drive robot with two traction wheels and two omni wheels, I think this is a pointless experiment. Whats the point of having 4 omni wheels if you don't move in omni directions. If you are going to add wheels, then its sounds like you're going to have 6WD, which has a much different dynamic for going over bumps then 4WD, so nothing from this experiment can be translated. If you are going to do some sort of holonomic drive, then you are going to be losing traction and again this experiment becomes useless.

As far as I understand, you don't LOSE traction with holonomic. However, you DON'T have as much power. In my opinion, this proves that you can make it up with the omnis. With holonomic drive, it'll just be slower.

[JesseK]

Thanks for posting this lynca. This confirms our K'Nex mockup's tendencies to auto-straighten the rear omni's after the front wheels crest the top. There was one instance where it didn't auto-straighten, but it looks like that was because the whole robot bounced a bit.

I'd recommend adding some weight to the setup, either via a small chield (j/k :D), weight room weights, or some other means. It'll be a whole different animal to control when there's mass preventing it from climing and/or assisting it down the other side.

[Andrew Schuetze]

Whats the point of having 4 omni wheels if you don't move in omni directions.

I am not a Discobot mentor however, I conjecture that the point of 4 wheel omni drive is to create as much turning ability from a tank steer drive-train as is possible on carpet. With just omniwheels on the front the robot pivots from the mid-point of the rear axle. You will notice in the video that this robot pivots from the center of the drive-train. Not a sweeping or gate turn but rather a point turn.

I too am interested in where, what type, and why they intend to add more wheels are posted by Andrew L.

Andrew S.:cool:

[artdutra04]

I am not a Discobot mentor however, I conjecture that the point of 4 wheel omni drive is to create as much turning ability from a tank steer drive-train as is possible on carpet. With just omniwheels on the front the robot pivots from the mid-point of the rear axle. You will notice in the video that this robot pivots from the center of the drive-train. Not a sweeping or gate turn but rather a point turn.

I too am interested in where, what type, and why they intend to add more wheels are posted by Andrew L.

Andrew S.:cool:You can achieve the same zero turn radius spin with two omni and two traction wheels if the omni/traction wheels are at opposite corners of the robot. So if the front left wheel was an omni, the rear left would be traction, the front right would be traction, and the rear right would be omni.

[Green2]

I noticed that the robot seems to fishtail at times. Is that due to operator or does the robot have some issues with launching and going straight at the start of motion?

[DTTrobotics1]

I find this experiment that you carried out to be very interesting.

We've been wondering of any way to actually get over that bump for awhile... I don't recall us testing that out yet...

The omni-wheels should also help with the carpet + rubber tire situation.

[GaryVoshol]

Have you considered adding the bumpers to the bot yet? If those are 8" omnis, it looks to me as if the frame member is 12 or 13 inches off the ground. The bottom of your bumper will be no more than 11". Bottoming out?

And a more general question which should be of interest for all teams - what will keep a ball from rolling more than 3" under the frame in this kind of setup?

[Brian Ha]

I couldn't view the video but from discussion from our team it would be very diffucult to get over bumps because of the vertical incline. Also watch that if you do put on omni wheels you would be pin pretty easyly(however you spell it) I could be wrong and maybe your robot could get over the bump but watch out for easy pinning.

[ke7qcl]

Have you considered adding the bumpers to the bot yet? If those are 8" omnis, it looks to me as if the frame member is 12 or 13 inches off the ground. The bottom of your bumper will be no more than 11". Bottoming out?

And a more general question which should be of interest for all teams - what will keep a ball from rolling more than 3" under the frame in this kind of setup?

If they are 8" omnis and the bumpers extend down about an inch it would still leave you with about 4" at the lowest point while crossing the bump, but you still would want to add a type of guard to prevent the ball from rolling under the robot. Some ideas might be either a chain that will flex when you go over the bump or a piece of pvc pipe.

[lynca]

I too am interested in where, what type, and why they intend to add more wheels are posted by Andrew L.


We plan to add more omni wheels to the front and back to achieve strafing.

We also have lots of testing/cacluations to do with bumpers and speed ratios.

Thanks for all the helpful ideas. We will keep posting videos & updates on twitter. Stay tuned ... http://twitter.com/discobots

[Andrew Schuetze]

We plan to add more omni wheels to the front and back to achieve strafing.

We also have lots of testing/cacluations to do with bumpers and speed ratios.

Thanks for all the helpful ideas. We will keep posting videos & updates on twitter. Stay tuned ... http://twitter.com/discobots

Interesting... in 2006 the Toltechs, 499, used just four omni wheels at 45 deg. in the corners to obtain the control you are talking about but we were not able to climb the diamond plate ramp with that set-up. So if I understand your post, you intend to set up a second drive train to give you lateral or side-to-side motion.

I am certian that you are working on dealing with the impacts that the front omni wheel will encounter when hitting the ramp at speed.

[ke7qcl]

you can either do that or instead use mechenam wheels which would be generally the same thing. i think that it would allow you to do the same but i think it would be easier to assemble because they need to be parallel and it would be easier to get over the bump than with 4 omnis at 45 degrees.

[Andrew Schuetze]

you can either do that or instead use mechenam wheels which would be generally the same thing. i think that it would allow you to do the same but i think it would be easier to assemble because they need to be parallel and it would be easier to get over the bump than with 4 omnis at 45 degrees.

Omnis at a 45 could not climb the diamond plate ramp in 2005. In fact, the plastic versions that where new from AndyMark.biz in 2006 shattered if we hit the ramp too hard. Mechanum wheels where not available from IFI until 2007.

APS

[KGood]

Just wondering, on Week 1 Pneumatics Test video where you showed the piston, what PSI were you using?

[lynca]

Just wondering, on Week 1 Pneumatics Test video where you showed the piston, what PSI were you using?

Our pneumatics test was at 60 PSI

We will try to post more videos of week 2 soon, stay tuned to our new website,
http://2010.discobots.org/node/16

[lynca]

I am certian that you are working on dealing with the impacts that the front omni wheel will encounter when hitting the ramp at speed.

Thanks for the warning. We have been testing omni wheel high speed impact without any failures yet, but we have our backup wheels ready just in case!

[falconmaster]

Thanks for the warning. We have been testing omni wheel high speed impact without any failures yet, but we have our backup wheels ready just in case!

video is not working.......

[DynamiteHack]

You guys've already built a robot?? We haven't finished brainstorming! :ahh:

OT: The more I think about it, the more it seems like using omni wheels in a simple skid-steer config would help this year; can you guys confirm that it skids/sticks less when turning?

[waialua359]

To add light on some of the points made in this thread.
Our team has already tested a holonomic omni wheel setup.
As far as going over the bump, the problem we encountered was the transmissions/CIMS bottoming out at the top of the bump.
Raising it or using stilts such as the KOP AM ones raises the COG considerably and that's the route we didn't want to take.

We did additional testing on the 6 wheel setup that you folks proposed and the driver's loved the way they were able to maneuver around the field.
However, using the FP is a bad idea for the drive and we decided to scrap the idea and go now to plan B.
Plan B meaning moving away from omnis and drivetrain setup.

[lynca]

video is not working.......

The videos are just links to youtube & vimeo. The website videos are a bit flaky sometimes. But the kids are working on improving the integrated video section (we usually don't go over 1000 hits) !

http://www.youtube.com/user/lamardiscobots

http://vimeo.com/user1102634/videos/sort:date