Crazy Drive Train Ideas

[Jared Russell]

You have to be careful that you think about this in the right way. There have been all sorts of crazy drive trains in FIRST. Ball drives. All sorts of swerves. Active and passive suspensions (especially this year). Walking file cards. Ball differentials. CVTs. 2 speed (AM-style), 3 speed (DeWalt + 222 ball lock shifter), 4 speed transmissions (33). Brakes. Triwheels. Traction wheels, inflatable wheels, omni wheels, mecanum wheels, treads. Nonadrive. The list grows every year.

Some of the above have been used to great effect on certain robots. Others tend not to do as well. What's the difference?

The teams that succeed on the field don't generally seek to innovate their drive systems for innovation's sake - they instead analyze the game and what they want their robot to do well and use that analysis to drive their choice of drive system. In other words, before trying to think of off the wall ideas, think of what possible situations would cause the "tried and true" FIRST drive trains to fall short.

Steps? Driving on sand? The return of regolith? What if you decide you need holonomic/swerve-like agility with extreme pushing power (especially if you decide that your team isn't a good fit for swerve, either financially or because of fabrication resources)? How can you ensure that nobody can cause your robot to turn if you don't want it to?

(I'm speaking about drive systems you would actually want to use on a competition robot here. In the offseason, or for demo purposes, sure, anything goes. But the teams that succeed on the field more often than not have competition on their minds before they go and prototype.)

[Bjenks548]

While I know that this is not the best idea as far as making a game applicable off season part, with summer coming and the GDC refusing to give me a game hint to work with, I am bored and need something to do. Please do not criticize what we are doing and keep the post to drive train ideas and discussion.

[JesseK]

To sum up the critics of this thread:
"An innovation needs to be set to a strategy, solve a specific problem and be practical before the design even starts."

Thank God DaVinci didn't think that way, nor did these guys
<Philosophy>
The methodology of complex innovation has been around since the Rennaisance Era, and didn't start with a specific problem to solve. The problems are abstract, even in the FRC Drive Train realm: design a drive train that is both agile and has high tractive power. The reality is that solving the problem is an excellent lesson in Innovation itself and the lessons learned will carry further than just the FRC season. The biggest lesson is that there is never a silver bullet to solve every problem in a mechanical system. Given that now is the Prototyping Season, this is a fun discussion with perfect timing.
</Philosophy>

I don't understand why all teams want the drive train to do everything for mobility. If the rules allow for it, why has no one even prototyped something like a tethered Quad Rotor that lifts off and gets the game object? Sure, the thing would probably get beat up during a finals match but the sheer success in prototype will spur a LOT of creative thought come build season. Of course just implementing a quad rotor with limited amounts of motors is a feat and requires its own innovations worthy of patents and PHD's, heh. Then there's the control component, which many programmers would salivate at the chance to do, and the sensor feedback component which boggles the mind of even the smartest professionals.

0.02

[Jared Russell]

Originally Posted by Bjenks548

While I know that this is not the best idea as far as making a game applicable off season part, with summer coming and the GDC refusing to give me a game hint to work with, I am bored and need something to do. Please do not criticize what we are doing and keep the post to drive train ideas and discussion.

I'm sorry if I came off as criticizing - I think that no matter what you decide to do, prototyping in the off season is very beneficial and a lot of fun! True innovation doesn't necessarily require a concrete problem first, but history shows that necessity is the mother of invention quite often.

I was just throwing out that if you want to prototype for fun and to for practice, why not also practice analyzing a FIRST-like game challenge. While the GDC hasn't given hints towards 2011 yet (and given past history, they are unlikely to do so until much closer to kickoff), history shows that elements of past games tend to be re-used in new and slightly different ways. Pick a game that you aren't familiar with, analyze some of the technical challenges, and think of ways to solve them.

Either way, you'll have fun and learn a lot.

[Andrew Schreiber]

Originally Posted by JesseK

To sum up the critics of this thread:
"An innovation needs to be set to a strategy, solve a specific problem and be practical before the design even starts."

An idea need not solve a specific problem but a design should. That problem can be completely off the wall or it can be realistic. A water game, not so realistic for FRC but figuring out how to navigate under water would be useful for NURC.

I suppose there are two ways of looking at this, either you can define a problem and develop a solution or you could define a solution and develop a problem. Personally, I prefer knowing my problem first as it means I can more easily break it down into sub problems to be solved.

As for crazy drive trains, why not do a definitive test 14 wheel drive (703) compared to 6wd or 8wd or 2wd? It might be that having more wheels has benefits that we can't see easily.

As far as problems you might want to look into solving I would suggest climbing stairs (6" or more) or navigating on a narrow area (see 2004 field centerpiece for example). These are challenges a lot of FRC teams have never faced and none of the students still on teams have faced.

[Bill Beatty]

Speaking about wild drives, does anyone remember and have any information on a wild drive I saw on a web cast a few years ago? It had three flats driven between two disks. I think the robot was decorated to look like a cow. I still haven't figured out just how that thing worked. A picture would be great.

Mr. Bill

[MrForbes]

Speaking of stairs, designing an FRC sized legged robot would be a great exercise, and if you can figure it out, might even work in a real game.

My son's friend Matt made a smaller one, and although the mechanics would probably have to be quite different (crankshafts/connecting rods for legs?) the programming could be similar. You never know where this kind of thing will lead.

http://www.technewsarizona.com/Artic...0091220&t=0600

[Alex Cormier]

Originally Posted by Bill Beatty

Speaking about wild drives, does anyone remember and have any information on a wild drive I saw on a web cast a few years ago? It had three flats driven between two disks. I think the robot was decorated to look like a cow. I still haven't figured out just how that thing worked. A picture would be great.

Mr. Bill

That would be team FRC 276. I had the pleasure of competing against them in 2003, 2004, and 2005 at Buckeye.

http://www.chiefdelphi.com/media/photos/15308

[Mr. Tech Vike]

Originally Posted by dodar

i think it would be cool to see a crab-drive tankdrive robot

my team did that this year we programed for a skid-steer and crab. all we had to do is make sure the wheels were lined up and in hae right direction (you can get turned around pretty easy) but all we had to do is hold down a button on our joy sticks.

[JamesCH95]

I've seen teams with two sets of tank drives oriented 90deg apart with the ability to switch between the two. That is an interesting problem to think through or build.

Can you make a robot that moved like an inch worm?

What about a tread design like The Pack Bot? Talk about being robust...

Have fun with your prototyping!

[Hawiian Cadder]

looking at the challenges up above i believe i have found a form of locomotion that would satisfy all of them.

a robot that uses Augers to contact the ground.


http://www.youtube.com/watch?v=1uynmApjhWI

better yet, a robot of this type would be somewhat holonomic if the augers were separately driven, it would also have incredible amounts of traction, even more than tank treads i believe. it could go up stairs if the auger diameter were large enough, and if the Augers were hollow or could float, it might contribute to a water game. i suspect that depending on the pitch of the augers, not much gear reduction would be needed for a working robot, a single AM stack-box would probably do the trick.

[EricH]

Build a drivetrain that can handle corn. You just might need it...

[jamie_1930]

Originally Posted by Hawiian Cadder

i just came up with another one, make a robot that is entirely encased in a hollow sphere with omni wheels on every corner facing outward, to drive the main part of the robot drives the inside of the sphere, if we ever get a water competition this would be highly helpful i think.

This is a drive train I've been pondering for a while and I would love to make it.

[CENTURION]

This is basically what I've been doing for the past few weeks straight

Specifically, I've been attempting to find a way of making a demo/promotional robot that our team can show off to schools, potential sponsors, and the general public (especially when we do recruiting for our summer camp).

I personally have been wanting to tackle the problem of an off-road sort of drivetrain, so I've been experimenting.

For now my favorite solution is off-road swerve/crab drive, using two wheels on each pod (ok, I'll admit, the two-wheels part is mostly to make it look cooler, but it would have it's benefits: traction, weight distribution on soft material, etc)
Each pod would be individually rotated via an FP or Window motor, and would raise and lower via a pneumatic piston and some guide bars. (the pistons could be pressurized somewhat to create a passive suspension system).

I will upload some sketches and better explanations later...

[548swimmer]

Originally Posted by jamie_1930

This is a drive train I've been pondering for a while and I would love to make it.

http://xkcd.com/413/

'nuff said.