[FTC]: Holonomic?

[krisloz]

I'm thinking of trying a holonomic drive next year and I just want to know how helpful and the problems teams faced in using this style. How hard was it to engineer/program? I'm going to use Robot C for programming. Any input you may have is greatly appreciated.

[JesseK]

Each year at least one holonomic-like drive makes it into Eliminations at champs. While the drive itself isn't the only factor in that success, the teams who drove those bots felt very comfortable with them. So get started with building and driving that drive train now.

There are several whitepapers here on CD that give example code of omni-drive, both driver and field-centric. Go to "CD-Media" at the top, then select "Search" under "Papers". Search for "holonomic", "omni drive", and even "Mecanum" since the code is all very very similar. Try not to dismiss any of the papers just because they're too technical -- they all have something of value.

[LeelandS]

I can't speak to Holonomic drive at the FTC level, but I can give you a little more insight on the Holonomic drive at the FRC level. I've never been part of a Holonomic project first-hand, so you may want to take what I say with a grain of salt.

I had a conversation a year or two ago with a team that was using Holonomic. Since I'd never seen the drive up-close before, I was pretty excited about it. The team told me the biggest issue they had on the programming end was the non-standard driving. That is, as he put it, "there's not really a way to just drive forward." That's obviously a result of the angling of the wheels, and none of them being pointed straight forward normally. As stated above, you'll be able to find a great amount of while papers on the subject of getting the wheels to work out well.

From the mechanical standpoint, I can't offer you much aside from what I've seen in pictures. On that note, I can say it doesn't LOOK terribly difficult. The biggest problem I've heard people have is not having the wheels lined up correctly (at 90 degree intervals). That's something you need to be wary of. Again, don't put a lot of stock in that. The mechanical difficulty wasn't something that came up much in the conversation.

The best advice I can offer is browse the media and white paper section, looking for examples of Holonomic drive systems and documents about how to program them.

[brennonbrimhall]

Originally Posted by LeelandS

I had a conversation a year or two ago with a team that was using Holonomic. Since I'd never seen the drive up-close before, I was pretty excited about it. The team told me the biggest issue they had on the programming end was the non-standard driving. That is, as he put it, "there's not really a way to just drive forward." That's obviously a result of the angling of the wheels, and none of them being pointed straight forward normally. As stated above, you'll be able to find a great amount of while papers on the subject of getting the wheels to work out well.

Keep in mind that I'm from FRC too, and my perspective is totally from FRC. Sorry.

FRC Team 1640 has a wonderful article about a non-holonomic, omni directional system. They have a nice focus on the driving part – though the ease of drive really depends on how you program and map your controls.

http://wiki.team1640.com/index.php?t...el_Pivot_Drive

Lastly, from a scouting standpoint at FRC, Holonomic drives are notorious for their low traction.

Hopefully this helps.

[RRLedford]

Having done two rather poor driving & underperforming non-tank schemes in the last two FTC seasons - one a 4-omniwheel drive with dual-wheel axles and one a Kiwi 3-wheel drive using 3-omniwheels per axle scheme (2nd bot was heavy), as head mentor, I was a little skeptical of doing anything but tank this season.

However, with the relaxed rules now allowing any 4" & under wheels to be used, combined with the low ~$50 cost of a (4-wheel) set of the VEX plastic hub mechanum wheels, this led our team captain to want to try one more time. I was still dubious, but allowed him to launch the experiment, as long as we went forward in parallel with a tank drive build too.

Well we just completed the DEC 7th Illinois regional in Chicago, and our team was the highest seed after qualifying rounds and captains of the winning alliance.

Though we had issues with our Block Party autonomous that kept having our bot become misdirected and hitting the wall, then driving under the balance beam and lifting/shifting the bridge, under threat of disqualification we reworked autonomous, abandoning our block dump route in favor of just a straight drive onto the bridge.

We were the best flag winders in the competition and by partnering up with one consistent autonomous I/R block scorer and two good end game hangers, our alliance went on to victory. The Mechanum drive was essential for the precise alignment of our flag winder shaft with the flag handle shaft, as many teams had adequate winders that only worked when they could properly position their tank drive bots, which most of the time they did not accomplish. Note how our flag winder scheme uses T-slotted aluminum framing to allow for quick and precise height adjustment of the winder shaft, as well as easy chain tensioning.

We used a non driven wheel axle made from a long 10-32 high strength bolt, that passes through a double brass sleeve bearing Tetrix channel overlap cube. This cube then mounts below our 20mm Bosch-Rexroth sloted aluminum frame base with a pair of 10-32 hardened screws and 10-32 T-nuts.

The tricky part was getting a 4-hole Tetrix hubs to integrate with the 7-spoke VEX plastic wheel hubs. Some strategically applied Shoe Goo between the two carries the main amount of the sprocket's torque over to the wheel.

Above the frame sits the motor with matching size sprocket sits located at a ~30º inboard offset from the wheel axle. The motors have adapter plates between them and the frame to allow for the needed lateral sprocket alignment offset. Motor adapter plates also mount to frame with 10-32 hardened steel bolts. This drive train is extremely durable and fairly precise (except when driving near 45º diagonally). We had good maneuverability on the bridge too, and stayed where we stopped. The power of the new Tetrix motors was very good at the 1:1 level.

This design allows for easy changes to wheelbase spread, as well as quick and precise chain tensioning. After competition, all four motor/wheel were still good enough to handle another day of competition with no tune up needed.
If you are tired of running with bent-axle tank drive schemes, I heartily recommend giving this directly driven mechanum wheels scheme a try.

NOTE: we did have to remove the illegal servo mount frames of the 1/4-scale servos on the morning of the competition, before inspection.

A few PICs of our drive train are below:

[F Elliott]

Thank you for posting such great photos. As a rookie coach coaching an all-rookie team who got their KOP two weeks ago, your photos are an invaluable resource. Thanks!

[RRLedford]

You are quite welcome.
Here is a PIC of our 2012-2013 Kiwi bot drive train



Later we added 3rd omniwheel per axle:



-Dick Ledford

[maths222]

Just so people know, AndyMark has released an FTC compatible 4inch mecannum wheel. It has the right hole pattern and has been designed to work with FTC robots.