pic: Seen at Fall Classic: Side

Side view of team 585’s offseason robot, competing as 9585 at the SCRRF Fall Classic. Drives smoothly, with two mecanums and a single omni. To say I haven’t seen that drivetrain before is an understatement.

1 Like

Wouldn’t there be balance issues with a drive configuration like this?
Certainly an interesting concept regardless; at the very least in certainly makes good use of space.

I didn’t see any issues. At least not in terms of tipping.

Wait, can it strafe? How do the wheels look while driving?

It can strafe. Can’t say I was paying attention to the wheels. Talking to the team, they started with mecanum code.

What’s the main body piece made of? It’s not fiberglass, is it?

Right the first time. You should see their regular-season robot.

Is that back omni wheel powered? It is hard to tell from the picture, but would make most sense if it was powered.

All wheels are powered.

Some of the chain runs actually drive the lift.


Seems like the U-drive concept is popping up all over this off-season (third I’ve seen posted, including my own), but this one definitely takes it to a new level, especially when it comes to controlling unintended torque, and also in wrapping around the game piece.

Are those chains or belts running to the back? Are the motors all the way back with the strafe wheel? That looks like too much slack.

I’m trying to understand the overall benefit of this configuration vs a standard mecanum setup or even front mecanum back 2 omni. Surely this config is not favorable in traction with only minimal benefit to powered rotation. Weight reduction (3 gearboxes vs 4)?

Three points of contact is deterministic at least and makes a suspension not needed to handle ramps such as this year properly.

Chains. Not all of them are for drive. I forget exactly where the team put their other drive motors–the outer ones look like the lift chains. (Of note: the midships gearboxes change chain direction to operate a lift, and are not part of the drivetrain. Unfortunately, they didn’t quite have the time to get that fully operational.)

Edit: Just for the record, I’m not affiliated with the team. But after seeing that… Well, here’s the result. I’ll have to see if I can signal one or two of the 585 folks to describe it in more detail. As I said in the description, I’ve never seen a drive like this before.

Yes, that’s part of it. It also reduces the amount of width and machining that would be needed to mount rotated omni wheels to produce a kiwi. The angles between the forces of adjacent wheels here are 90- 135-135 as opposed to 120-120-120 for kiwi, but it should be close enough to work after some tweaks to the inverse kinematics.

A peculiar thing I just noticed after mentioning kinematics is that the robot can deliver more drive force in the direction transverse to the direction it picks up totes than in the direction parallel to the totes, by a factor of about 1.7 for equal loading on each wheel, more or less as the CoG moves. This means that when you want to cover ground, you’ll want to think of the long edges as the front/back, and the open and closed short edges as the sides.

Edit: If this drive style does not yet have a name, I dub it Meckiwi.

I noticed that there are manipulators on the robot. How did it perform at the event? were there any break downs and if so, what were they?

Honestly, I don’t remember much, particularly as far as breakdowns. (Though there was that one wire that needed to be taped into place…) Performance was decent, not spectacular–about what a kitbot could do. I don’t recall seeing the lift lift anything, though it did at one point snag a tote.

Anyone have video of it in action?

It would be interesting to see video of the drive in action, especially considering it didn’t have to deal with a changing center of gravity w.r.t. totes. A closed frame design may alleviate some of the challenges with the long chain runs.

For a game with exclusion/safe zones, this setup could be interesting for specific robots - like 2012/2013, if you discount the balancing in 2012. Or even perhaps for a 2014 robot with a specific team/strategy mindset - after all, the Killer Bees proved all-omni was plausible in a defense-heavy game.

I’m all for reducing the number of wheels required to play :smiley:

Off topic, but what 33 actually demonstrated is that floating like a butterfly can be a winning tactic, IF you can also sting like a bee.* This requires a very good driver. Which they had.

Muhammad Ali demonstrated the same thing 40 years earlier, in a different sport.

Doesn’t look like it, but I do know there was a stream, and I’m pretty sure the stream was saved. It’s just that finding it is going to be interesting.