"How does it drive?!?!" Team 857's bot

[GregT]

This kind of idea died quickly with my team. We prefer to hit our problems head on, so being able to get out of sticky situations isn't as important to us as being able to overpower grandma's robot (im sure i'll eat these words ).

Cool drive system though,
Greg

[patrickrd]

Awesome drive system!

As Greg mentioned, our team considered this very same type of drive system (which is called the omnidrive in the Ithaca area). Our Cornell faculty advisor was a strong advocate for that same drive system. He also happens to be the advisor of Cornell's Robocup team (3 time international champions), who came up with that type of drive system (on a much smaller level) several years ago.

Ultimately, we decided that the 50% loss of power to spinning the rollers in the wheels was too much to give up in this year's competition for increased maneuvarability.

Either way, i'm glad to see it successfully implemented and, depending on your strategy, you could have a huge advantage over other teams (not to mention the design awards you deserve).

- Patrick

[ajlapp]

Thank you to everyone who has had kind words about our new drivetrain! We've spent two years developing and testing prototypes.....finally FIRST came through with three matched motors and a flat field.

Everyone who is coming to the Chicago and Western Michigan regionals will get a chance to drive one of the Killough Chassis'(we'll have a working model there) and see the original spherical wheeled design.

Keep an eye out for the controls.......they are the real spectacular part!

[ajlapp]

Surprisingly enough you really can't move the thing sideways. In order to move it you still have to backdrive at least one motor. It's certainly not designed to be a pusher, but traction is not at all terrible.

It's fast and nimble, but can be confusing to drive if you lose your orientation....we've been using a colored flag system for the drivers!

If you've got more questions please post them......we built it for you guys. We love to toot our horns about it.

Hey Joe, any comments? Brian? Andy?..........does it stack up against the swerve?

See you all soon. And let's see some bots.

Anthony Lapp
Team 221------> now 857

[Mike Soukup]

Originally posted by ajlapp
Everyone who is coming to the Chicago and Western Michigan regionals will get a chance to drive one of the Killough Chassis'(we'll have a working model there) and see the original spherical wheeled design.

I'll be first in line to take it for a spin. I'll try to take it easy, but the only speed I know how to drive at is fast

Mike

[Adrian Wong]

How did you guys finally conquer your control system? The last I heard, you guys were making a custom control stick. But your post said something about hideous equations in PBASIC. Did you end up going with a standard joystick and doing some fancy math?

[ajlapp]

We won't tell just yet................it's our only team secret(just for fun)

We did actually conquer both systems though................we leave up to the drive team which device they ultimately prefer.........I will say that we were able to build a system with no feedback no sensors and no complicated machining.

Talk to me after the regional or during if you'll be there and you can drive the bot and see the final solution!

Anthony Lapp
team 221--------> now team 857

[Matt Attallah]

How did you get it to go straight, if it does. It seems at that angle of the 2 other wheels, just 1 wouldn't be strong enough, and thoes disks that are in the wheels, that would give you great pushing ability, but, agian, how can you go straight? I haven't watched the video and, no offence, don't plan to due to me being confined to my 56K!!! Mabye i can watch it in school...

[Adam Krajewski]

To drive at a linear velocity V in the direction of theta from any vector starting at the reference point Xref, Yref with an angular velocity around Xref, Yref of psi, the wheels (radius of R, and length L1, L2, and L3 from Xref, Yref) must have a rotational velocity of w1, w2, w3.

w1 = (|V|/2R)*(sin theta - sqr(3)*cos theta)+(psi*L1/R)
w2 = -(|V|/R)*(sin theta)+(psi*L2/R)
w3 = (|V|/2R)*(sin theta + sqr(3)*cos theta)+(psi*L3/R)

Just calculate those and you can drive in any direction you want, while spinning, including "straight". The omnidirectional holonomic nature really creates no "front" or "back". We arbitrarily chose those sides. In the case of the prototype drive, the 'front' was perpendicular to the wheel driven by the Fisher Price motor.

Adam

[l1jmx]

I remember reading something about the original prototype being started almost two years ago. Did you guys branch off of another team or were you working on this for some other robotics project? Your team number indicates rookies, but being another rookie team I'm simply in awe of what you guys have done. Our little 4 wheel, 2 motor traditional driving thingamajig pales in comparison. In any case, congratulations, and I'm sorry I'm stuck over here on the west coast 'cause I'd love to see it in action.

[ajlapp]

You caught onto our scheme........we're not actually a rookie team.

This year due to some sponsor changes and other political stuff, we had to change our number. This has technically made us a rookie team, but some of us on the team have been doing this for as long as 7 years.

We've been working on the drivetrain for about two years, and we finally got lucky enough to get a game that would allow us to use the drivetrain. The flat field and three matched motors helped a lot.

Hopefully we'll have a chance to see each other at Nationals.(fingers crossed) We're really hoping for a technical award for our work on the drivetrain and it's controls.

Anthony Lapp
team 221 -------> now 857

[l1jmx]

Perhaps we'll see it if the nationals are televised or such. In spite of our being an even numbered team, we really have no plans of attending nationals (more for monetary reasons than anything else - raising enough to register for regionals AND nationals as a real first year team was beyond our ability). From what I've seen though, I'm rooting for you all for some kind of technical award - I love seeing cool devices like that and would have loved to have worked on building it. Since next year Nationals will likely be just for odds (and it's unlikely or team will be a top scorer at regionals this year), you probably won't see 840 until 2004 (at which point I'll probably be a mentor on some other complete rookie team started up by whichever college I end up attending). Maybe I'll make it there some day, but best of luck to you and your team. Keep up the good work and the cool innovations.

[Robert Flanagan]

we are currently trying to program an omnidrive vex bot with three wheels. even after looking at all this, we are confused. an elaboration or even better, the actual code would make us very happy. oh, and we are using big boy programming, not easyc

[EHaskins]

First you should check how old the thread is. This one is 4 years old! I don't mind when people bring back an old thread, but some people here think you shouldn't restart old threads.

For holonomic drive info check these threads.
Holonomic Drive Math
Vex Holonomic w/code in post#8

If you have any other questions PM me.