Holonomic drive and ramp

[robotaddict]

Okay, someone tell me when I go crazy. Here goes.

I was thinking that a holonomic drive bot could make it up the ramp with some attention to design and ground clearance, but my team disagrees. EVERYONE, including the people who support holonomic drive, thinks that it lacks traction enough to haul itself up the ramp. I also think that the major disadvantage for a holonomic bot attempting this is the contact surfaces distance from the ends of the 'bot, but on closer inspection I noticed a way to overcome this: tilting the wheels.

Because of the perpendicular rollers on omniwheels, the contact surface of a tilted wheel, say 10 degrees from horizontal, is the same as an upright wheel. All that's required is some fancy-pants mounting of motors and what-not. If the wheels were tilted in to 80 degrees from vertical, the contact patches would be closer to the corners of the robot that almost any other design.

With these tilted wheels, you could lower them from beneath the frame by enough to start climbing the ramp and get to the crest. However, once at the crest it is not necessary to get your entire robot up onto the platform if you can get your front wheels over the egde to sort of "hook" you on. Not on the platform, but more importantly not on the carpet. 5 pts.

If you could tell me what you think of this I could stop doubting the species of the mushrooms on that pizza I had. It would be most helpful!

Our team is a third-year team with, at the moment, limited funds, no machine shop, and some people who seem fine with two-wheel drive. If anyone more knowledgeable than me (head programmer w/virtually no experience but a good head for math & spacial relations) would put their two-cents in it would help our team do something interesting this year. I also think it would just be really freaking cool.

Anywho, I'm rambling, so I'm just going to cut myself off.

P.S., one of the few things my team has agreed on is the ability to climb the ramp, but seems to be putting this over our agreed upon chief priority of a freakishly maneuverable shooter, hence the holonomic drive idea. Ok, bye for real this time.

[mechanicalbrain]

Is this robot going to be the standard square framed omni system. If so I'm thinking your issue won't be traction but center of balance. Remember if your center of gravity is past the line of your point of contact it will tip.

[Ianworld]

The only issue with this sort of design, besides the obvious dificulty of tilting your wheel mountings is that you won't know for sure until you build the robot. Its very, very, risky to go into such a complex task thinking that it will maybe, just possibly get up if we just hook the top. I'm not saying that it wouldn't work, just that I don't know and that nobody will for sure until you build it.

So think about the game and decide how necessary getting up that ramp is for your strategy. If its not that important than maybe this is something you could try.

Of course the actual construction once again would be really, really hard.

[Not2B]

You'll want to work up a list of robot priorites.

Sounds like you don't have alot of machining and money... maybe keeping it simple is a top priority. Maybe simple is more important than freakishly manuverable. After all, a cool drive that doesn't work isn't as fun as a simple drive that runs every match.

There are lots of different ways to work up a decission process. Best bet would be to find one and use it - and stick with the answer even if it's not what YOU (or any one person) wants.

Example:
For our drive, we ranked it based on the following items, in order of importance (in other words, 1 is more important than 5)
1) SIMPLE
2) RELIABLE
3) Controlablity
4) Torque
5) Stablity

Speed, manuverable, ground clearance didn't make the top 5, but they were also important.

Crab, Holonomic, 2wd, 4wd, 6wd, tracked, halftrack, etc... were all ranked in each of the top 5... we had a tie between 4wd and 6wd... we voted, tied again, decussed, used facts and numbers, and selected our 4wd, mainly because we KNOW it's simple and reliable because we have done it before. And even though we want to upgrade to a 6wd, we think we should practice in the summer, something our budget hasn't allowed yet. (I wanted 6wd, but I lost, and I'm OK with that.)

Good luck!!!

[Alan Anderson]

Originally Posted by robotaddict

...tilting the wheels.

No, you're not crazy. I saw that idea used last year on Team 116's robot.

[Alekat]

I just have to say that sounds so wild. And especially seeing that picture of 116's bot i get what you mean. Simply an awesome idea.

[JamesBrown]

the only problem I see is slippage, if the rollers have a decent abmount of grip with the diamond plate you should be fine.

[Lil' Lavery]

116 basically had the design you are talking about last year. Thing is, it definately cannot climb. It's not because of slippage, but rather that the drive wheels will lose contact with the floor once your robot becomes inclined.
http://invisiblerobot.com/robotics/r.../p1060294.html

For example:
Pretend your driving your "north corner" (or face depending on where you mount the wheels, it doesnt matter for the purposes of this occurance) first. Once it hits the ramp, it will be pushed upwards, but the south corner, not having hit the ramp yet, will still be on the ground. The robot as a whole now has a few degree tilt on it. Because of the tilt, the east and west corners are slightly lifted off the ground, and after you drive a bit more, will completly lose any contact with it. Unless your robot is perfectly balanced, the robot will now tip twoards either the east or west corner, returning that corner to contact with the floor, but removing the other corner further from it. With the absense of this corner, you cannot drive straight forwards anymore.

Our holonomic drive even expierienced the beginnings of this problem with the mere climbing of the triangular loading zones last year.

[sanddrag]

I'd say it will definitely have trouble with the ramp for the abovementioned reasons. However, you could always add a pneumatically actuated drop down assistance mechanism (wheel, wheels, or roller) to aid in the climbing. You could even power it with something like globe or window motors or fisher price motor(s) and gearbox.

If I had my way, this is what our robot would be like this year. I think omni drive is perfect for this game (crab might be better but it is more complex) But it is not my robot, it is the students' robot, so we will be having a skid steer (tank) drive system.

[kaszeta]

I agree with sanddrag, there's no particular reason your bot can't be both holonomic and handle ramps. There are fairly easy ways to accomplish this, either by tweaking some of the details of the holonomic drive itself, or have some sort of mechanism

I think holonomic drives are going to be popular this year, since it's an attractive approach to the problem. Heck, I think the students on my team have thought about it, even.

Test early, however. Holonomic drives can do some weird things.

[Lil' Lavery]

well, you could build various options to assisst with the ramp climbing, such as an active suspension etc etc etc
But by doing that you make the whole drive system about 2-5x as complex, costly, heavier, and most power absorbing. It may be worth it, just depends on the rest of your robot.

[jerry w]

Originally Posted by robotaddict

P.S., one of the few things my team has agreed on is the ability to climb the ramp, but seems to be putting this over our agreed upon chief priority of a freakishly maneuverable shooter, hence the holonomic drive idea. Ok, bye for real this time.

i think that you are looking at the problem with the wrong perspective. here is how you gain the platform.
1. accelerate the robot to full speed.
2. at the moment that you reach the ramp, reverse motor power.
3. most robots will, at this point, topple forward onto the platform.
4. if robot momentum was not enough to remove everything from the carpet, then an alliance partner can scoot you up the ramp.

when i first heard this proposed, i laughed. but with a shock absorbing material on the front of the robot, it cushions the fall and helps to stick the robot to the ramp.

jerry w