pic: 3 wheel omni drive concept



This seems like a solid idea. I can’t foresee any major issues with this type of drivetrain.

Your avatar seems very suited to this kind of drive.

In all seriousness though, what exactly am I looking at here?

Love the kickstands.

First make a segway robot. Then add a third wheel so it can go sideways. Your robot’s profile is now 1 foot wide when it sprints down the field.

The self righting mechanism is probably more entertaining, but the kickstands are an absolute necessity to start a match. Building them this way lets you lock them in a slightly over center configuration when deployed, so the robot won’t tip even without air.

If this ever makes the competition field, I will personally buy you your own personal Segway.

He’ll call you on this in 2025 or something when things have changed so much that this is viable.

What’s the piston on the side pushing the strange half-gears doing? I can see it pushing the two little wooden paddle things, but those don’t seem to be attached to anything.:confused:

Kickstands so it won’t fall over pre-match.

Can you make a gif of the kickstand and gear motion. I am having a hard visualizing the motion.

Is the self-righting mechanism pictured?

Interesting!

Is the strafe (center) wheel actuated to provide a constant (or adjustable) force against the carpet? As the robot leans into “forward” acceleration, the strafe wheel will otherwise tend to rotate off the carpet.

What is the “peg” sticking out the side just to the left of the strafe wheel?

If hit hard enough on the long side, or if power fails for a moment, it will go down. If I were to do something like this, I would use a kickstand that could be used to pick the chassis up off the carpet afterwards. (sniped!)

With this extreme an aspect ratio, it is possible (under 2016 rules) to have a gap in the bumpers so wide that another robot can hit your robot square and hit nothing but frame perimeter. I would anticipate a change in bumper rules after the first year this concept hit the carpet.

The rules state 8 inches from each corner. This is to stop metal-to-metal contact. If this robot “strafed” between another robot’s bumpers, it would be metal-to-bumper contact.

Also, of course, nobody who is aspiring to win much more than Innovation in Controls will make this, so I don’t think the GDC needs to panic yet.

Judging by the appearances and the finger joints, it looks like a majority of the frame is made out of wood. I would be a little sceptical of bumpers hitting wood repeatedly in high-speed collisions.

Other than the clearly ridiculous nature of this robot, the main thing I see being a real hiccup with it is the shape of the center omniwheel. Since it isn’t really round in profile, I don’t think the robot will tip forward or backward on the wheel smoothly. It’ll take weight off of the left and right wheels when the robot tips at certain angles and put more weight on them when tipped at other angles, and it may prevent the robot from righting fully at certain points. I’m not sure I’m describing this well, might need to draw something.

I think this could be fixed either with an active suspension, or in a far more boring way by making that third wheel outboard of the other two, forming a sort of kiwi drive. But the latter would destroy the whole concept of the robot, so where’s the fun in that.

This is far from the first wooden robot frame in FRC - teams have been doing laser cut wood frames for years. They can be done very robustly.

The middle wheel is sprung. The module pivots around the peg that you see and is forced into the ground with a spring (latex). This keeps it’s downforce constant into the carpet. I had an idea for an adjustable force mechanism, so you could prioritize whether the center wheel or the outside wheels saw more load depending on which direction you were accelerating. It seemed a little farfetched though.

This is what the robot would look like with bumpers. The kickstand is needed because the self righting mechanism is really large and cannot be deployed at the start of the match.

http://i.imgur.com/gzhdFvV.png

And here’s another look at the kickstand mechanism:

http://i.imgur.com/JLSYKUE.png](http://i.imgur.com/m435sxC.png)

Wood is good! It’s far and away my favorite material. With bumper rules in place, you don’t have a lot of metal on wood contact. I wouldn’t expect anyone to hit this type of robot very hard anyway, because there’s a chance it would fall onto your robot if you get too rough.

Um… what about high centering??

I wasn’t saying that wood robots won’t work. I’ve seen a few of them and they often work well. All I was saying was if there is a gap in the bumpers large enough for another robot to hit inside, you could see legal wood on metal contact (with a manipulator outside the hitting robot’s frame perimeter) and that scares me.

As for not wanting to hit that robot because of the possibility of tipping it: if I were a ref and I saw a robot play defense on yours resulting in a tip, I would be hesitant to call that an intentional tip. When you build a robot like that, you should expect heavy defense and make very sure your anti-tipping mechanism work as well as the inherent anti-tipping properties of a normal robot. Teams should not penalized because they were defending against a robot that isn’t well build to play the game (which includes defense).

I think the tipping thing was moreso I don’t want a robot to land on top of my own, rather than the tipping penalty thing.

Why not use larger omnis for the sides? You could get your CG below the axles and then you wouldn’t need nearly as complex mechanisms or software.

There isn’t enough of a lever arm to a CG below the axles (even w/ 8" wheels) to keep the robot stable when accelerating.

It also just isn’t nearly as cool.