Octocanum vs. Tank

Hi CD,

A teammate and I are talking about drivetrains that we want to play around with and figure out over the summer, and there is some debate over which drivetrain is better. He is arguing for octocanum, and I am arguing for a tank/WCD drivetrain.

He is saying that:

  • If done well, octocanum has more maneuverability
  • If done well, octocanum (with one CIM/mini CIM per module) has more power
  • We can do more than just tank
  • Octocanum adds to the overall robot performance
  • It’s more fun to build octocanum
  • We have all summer to build it and figure it out, so why not?
  • If we can make it and it works, it would be the better option

I am arguing that:

  • We have never used pneumatics before
  • We have never made custom gearboxes before
  • We don’t even have a sure means of machining yet
  • If done correctly, tank drive will not prevent you from performing well
  • Tank drive is much less prone to failure
  • Tank drive is lighter
  • Tank drive is simpler than octocanum
  • Octocanum would take up a good chunk of our CAW (we don’t even have Talon SRX/encoders yet)

The octocanum modules he’s considering look something like this (and mount to bearing holes in VersaFrame): https://www.chiefdelphi.com/media/photos/45303? OR Versadrop with dual motor COTS gearboxes of some sort
The tank drive I am envisioning would be a 6 CIM 6wd mostly COTS design. Something like WCP SS gearboxes/VEX ball shifters and VersaFrame

For Steamworks we used 4WD mecanum

Any advice for which drivetrain we should spend our time putting together and figuring out over the summer?

Side note: We both agree that a drivetrain is only one piece of the puzzle.

You’ve got to walk before you can run. I’d say, ensure you are capable of building a simple and solid WestCoast/Tank drive before exploring other avenues. If you have the means, do both! But otherwise, make sure your team is can design and build a proper tank drive.

It really depends what your high level goals are. If your main goal is to grow your skill set, then Octocanum will certainly do that, as there is a good chance that it won’t work very well in the first iteration. If your main goal is success on the competition field, then I think the answer is hands-down tank drive. It is likely that both of those are goals, it’s up to you to figure out which is more productive.

Talking from a purely speculative perspective, I think a second year team should hold off on Octocanum and perfect a good tank drive (probably WCD) first. It looks like you are falling into the trap of getting far too ambitious for your resources. Still, you know your own resources much better than I do.

In general I have a hard time finding a game where the added complexity of octocanum doesn’t make it a worse choice than a solid tank drive. Tank drives win tons of competitions year after year. I also feel that if you have the mechanical chops to pull off octocanum, you might be able to get a better educational value of doing a swerve drive.

You don’t think Octocanum has advantages over regular tank in that it would most likely have more power (in terms of pushing), and more maneuverability? Both of these things are crucial to winning competitions, don’t you think?

Let’s look at the negatives to octocanum: it’s heavy, it’s complex, and it takes up more space. You can negate those first two with a great design, but is it really worth it? In regards to pushing power, you have 4 + (2/3)4 Cims (=6.67). Is the added complexity worth 2/3 of a Cim, especially since the use of a mini-Cim will hurt the efficiency.

My 2 cents, but a perfected 2 speed West Coast Drive with chain in tube will be useful any given year. Rarely is an increase in maneuverability worth the added complexity and time spent designing and building such a hybrid drivetrain. If you want to gain design and building experience, octocanum can be a fun project. So can designing and building a Swerve drive. If you want something to give valuable experience for next year with near total certainty, designing a lightweight WCD will be very valuable to your team.

Hereis one of our current designs. If we were to go with octocanum, we would do something like this, however, it definitely does need a lot of refinement still. We calculated the weight of this drivetrain (4 pods + frame) at around 35lbs, which I believe is an average (if not slightly heavy) drivetrain.

There’s no reason that an octocanum would have more pushing power than a tank.

Whether the additional maneuverability is “crucial to winning competitions” can’t be answered in a vacuum (254 does a tank drive every single year, and I’d say they win plenty of competitions). It can be an asset, certainly - but it can also be unnecessary. It depends (among other things) on the game, your robot, your strategy, and your driver.

I don’t even know if octanum beats tank at pushing power and manuverability. You can still run a 4 mini-cim + 4 cim drivetrain with tank but few teams do. Just because you have more motors doesn’t mean you have more power; you’re still limited by what an FRC legal battery can give you. Something else to consider is the fact that when you get in a head-on pushing match, your front wheels will get lifted into the air and you lose half your power. Also, if you designed your modules poorly, you might get lifted back onto your mecanum wheels, and we all know how well they do in pushing matches…

IMO tank beats beats mecanum at maneuverability. Normal wheels are just better at acceleration while moving forward and turning. These aspects of a drivetrain are more important, especially considering “moving sideways is a wast of time.”

Do 4 Cims and 4 Mini Cims not overpower 4 or 6 Cims? (4*.66) + (41) > 61, assuming thats how that works.

Any drive designed with pushing as a concern is going to be traction-limited, so additional motor power won’t make any difference.

(Additionally, that’s not quite how the motor power math works, regardless because the battery isn’t a perfect voltage source - but it doesn’t matter in this case anyway.)

Make sure your team knows how to build a good WCD before you tackle anything more complex. There’s a reason why people always emphasize it on this forum. It’s a design that is applicable to almost any FRC game in recent history, mechanically simple (thus less prone to failure), and highly versatile. Keep your robot simple, and you will reap the rewards of competition and gratitude from your pit crew.

Anyways, if your team really wants to build an octocanum drivetrain in the future, I’d start off by trying your hand at building a COTS shifting WCD over the summer. You’ll have an opportunity to take apart and assemble gearboxes engineered specifically for FRC use designed by engineers that have been in FRC for a long time (I recommend taking a look at options offered by VexPro or WCP). Additionally, the pancake pistons that a lot of shifting designs utilize are the same kind that teams often use for the octocanum modules. Familiarize yourself with your motor controllers and encoders. The experience you’ll gain will help your team’s foray into more complex drivetrains in the future much easier and less frustrating.

An important thing about octocanum (as well as all complex drivetrains in general) is that your drivers will require practice in order to use it effectively. Just teams that run a regular mecanum setup often forget that they have strafing capability and drive their robots like a tractionless 4wd tank.

You don’t think Octocanum has advantages over regular tank in that it would most likely have more power (in terms of pushing), and more maneuverability?

When it comes to pushing power its important to remember that gearing and the amount of motors on your drivetrain play an important role. Certainly, a 4CIM WCD geared way too high for its own good will get pushed around by a more conservatively geared octocanum drive, but most WCDs you’ll encounter in competition will either shift down to a lower gearing that’ll allow them to push you or shift to high gear and maneuver around you.

Octocanum has the option of either having mecanum geared low for precise scoring and traction high or traction low for pushing matches and mecanum high for maneuvering. In both cases its very important to be careful how fast your high gear is, because in the case of traction you’ll be sacrificing pushing power and in the case of mecanum you’ll be sacrificing maneuverability (mecanum drives geared higher than 15ft/s tend to have trouble strafing properly).

Keep these points in mind when comparing octocanum to tank drives. Do your research and best of luck with your summer projects!

These are all very good reasons to do octocanum over the summer because you have a lot of time to mess with it and you will learn more from it. Learning things like pneumatics, custom gearboxes, and getting more comfortable with your machining will help you with other areas of the robot, not just the drive. A good WCD can be designed during the season pretty painlessly so even if you never get the octocanum working you learned something from it and will still be fine for the season.

I will opine that this can be really bad advice, depending on the team.

A WCD is simple, but still far from trivial, and if you’ve never done one before it’s likely well worth gaining experience building one in the offseason.

You know, about that whole maneuverability thing…

Go take a look at 254’s robots again. Look at 1678. They are maneuverable, and very fast, and can push. And they can’t move sideways on their own.

Yep, tank drive.

It’s not the drive you use, it’s how you use it.

As a second-year team, I would build a WCD/mecanum drivetrain. NOT octocanum! Seriously, either one can take time to get right, why mix elements of both and have a really lousy time. So build a frame that can handle both (it’s possible) and start populating with WCD. If you have time at the end of the offseason, switch it to mecanum. Now you have the tank part, and you have the mecanum part. Next offseason, work octocanum if you’ve finished the mecanum.

I should note that both WCD and mecanum are practically “pre-packaged” drivetrains–you can go COTS for most elements of both–the hard part is going to be running the frame so it can handle both (er, not hard at all, if you use a couple of tricks like having extra gearbox mounts built in to start with). Then you’ve got most of the equipment to step up to octocanum.

Why do you say moving sideways is a waste of time?

Just playing devil’s advocate… what if time isn’t a concern?

Because 1986 and 3310 lost to 254?

“Work expands to fill the time allotted.”

If time isn’t a concern, then I’d DEFINITELY take the time to do a deep dive into WDC and mecanum. I know of a team that took two full offseasons to develop a mecanum drivetrain, and *still *hasn’t used mecanum (over a decade–and many highly competitive events and blue banners–later).

Here’s the thinking behind doing WCD and mecanum first.

  1. You need elements of both. WCD’s tank drive and pneumatics go towards the gearboxes and orientation changes of an octocanum, while mecanum programming and familiarity is required by said octocanum.
  2. Both are COTS. You can buy the parts, assemble them, and play with them. This would actually enable you to get to octocanum faster. (Partly because you’ve already got some of the parts.)
  3. Octocanum is what most folks would call an “advanced” drivetrain. It’s complicated. WCD and mecanum are “moderate”–they’re more complicated than the KOP drivetrain (“simple”) in that there are more moving pieces to deal with, but they aren’t as nasty as, say, a swerve to put together and have operational.

Long story short, if time isn’t a concern, you have more time to do more in-depth on the basics before trying something that you’ve never done and is something that very few teams do even now.

While I agree that teams should focus on scoring points instead of making complex drivetrains, 1986 and 3310 most certainly did not lose because of their drivetrains.

They beat a lot of other teams that couldn’t move sideways to get that far, my own included. (Also just a note, 3310 isn’t omnidirectional, but their partner 3719 is.)