To address 2018 specifically: I would not consider that field to be open enough for a holonomic drive to shine, regardless of team constraints.
Omni-directional drives do not require open fields to shine. Look at the applications in which omni-directional drives are used in industry. Typically it’s confined spaces that require precision movement. Things like warehouses and industrial plants. The ability to make fine adjustments laterally without having to rotate or move forwards/reverse can provide value in confined spaces.
5 Likes
100% do a kitbot in a weekend, make as much of a practice setup as you can, and drive it like to stole it for 6+ weeks. Drive, break, repair & reinforce, drive, maintain, drive, drive, break, fix, drive again. Heck, build another kitbot if you can as a testbed mechanism development. If you’re looking for sophomore on-field mobility by reading this thread, there’s no question. Drive the kitbot.
It’s not a year; it’s a season: off-season. If they know it’s a 2018-type challenge, they’re asking what new drivetrain they’ve never built before they should build after kickoff. If they’re asking that question and they care about on-field performance, the answer is kitbot. Every year.
3 Likes
This is a good point. My own team did drive systems in 2013 and 2015 which were omni-directional not for reasons of being able to get around the field fast. Those robots effectively spent all their time in protected zones - like an industrial use.
Assuming that we are talking about something like X-drive or 3-wheeled holonomic or mecanum where you are trading off pushing power, still I don’t think I would endorse that for the 2018 game though. Precision only mattered at the level of competition where the scale was getting very full.
The offseason after they feel like they have a capable understanding of how to design and build the structures above a DT. Teams should run a simple drivetrain that is bulletproof until they know what their capabilities are for the game to game designs.
Figure out how/or where to get a lift or a pickup or a shooter. Knowledge is key!
1 Like
This was my 16th year competing as part of 1339, and we have only run omnidirectional drive in-season once, in 2015, with mecanum. Not saying swerve isn’t superior to tank, but we’ve done okay (and this year particularly), with WCD. Our team even got to play with yours in Orange County, where our WCD and your swerve were a good partnership. Until we have another game like 2015 I wouldn’t consider X-drive at all, and even then it would be lower on my list of options. As others have pointed out, this style of drive has been a known quantity for many years and doesn’t get used in FRC for good reasons.
Thank guys very insightful knowledge hopefully this thread helps inspire any knew teams out there with what direction to pursue during their offseason and in season.
imagine someone invents a suspension swerve so it can go over obstacles lol
it was definitely very fun forming alliance with you guys and playing together. It was epic seeing wcd and swerve work together also beautiful robot guys we are def trying to get to your guys level of aesthetics lol
Definitely check out 4143’s swerve in 2016, 1533’s Swank drive from 2016, and 2767’s mini-suspension on their 2017 swerve. The last of those I question the efficacy of, but it certainly didn’t hurt their performance that year.
Interesting point - this (or the three wheel variant) can simulate swerve (so can mecanums but the you end up slower)
It requires simulating the turning rate of your wheels but because there are less constraints on omni wheel platforms than swerve (with them being more truly holonomic) it can easily be done.
I would go with west coast drive for the following reasons.
- if your team is strong enough you are going to get defence played against you, even if you are not the strongest robot in the alliance, it will be very easy to shut you down with omni wheels.
- if your team isn’t strong enough and your goal is to be a 3rd robot in alliance slection you might end up playing defense and you can’t do it well with omni drive, my team wont even think about picking a robot with omni wheels to play defense.
the only omni wheels robot that suits 2018 in my opinion is a robot that it’s all purpose is getting cubes outside of the field for power ups, but you don’t know for sure that you will get picked for this purpose only.
2 Likes
I’ma just leave this here…
While I haven’t driven a mecanum bot, I have driven a full-FRC scale X drive bot as well as modern swerves. The difference in traction causes a very noticeable difference in driving characteristics, to the point that I’d guess it’s actually closer to mecanum than swerve (if not nearly identical).
The VEX bot you’re driving is incredibly light weight, and therefore has a lot less inertia. That is likely why it felt and looks so snappy. My experience with a ~150lb robot was not the same (though this was in the age of CIM motors).
So what I’m seeing is that if you gear down your omni drive by a factor of sqrt(2), you will have the exact same kinematics and force of a mecanum drive. Interesting. They should behave very similarly.
Not quite; look at how the speeds and forces vary by direction.
All the kinematics are the same if you divide the omni by sqrt(2).
All the forces are the same if you multiply the omni by sqrt(2).
Admittedly, I’m sure there’s other practical differences between the drivetrains. But just based on those equations they look like the same drivetrain with one geared faster.
Ah yeah, my bad, I misread the diagonal column.
In practice, fully-weighted X-drive does not actually reach that full speed due to roller friction. The performance of all of these drives once you take friction into account is…rather different.
The best way to get a feel for how they handle is to build one.
Uneven contact with all 4 wheels is a HUGE problem! This means not only must frame be level, but surface you are driving on must be level. This is not EVEN taking into account obstacles which will GREATLY affect the force vectors being applied in the correct direction. Three wheel omni in a triangle configuration will eliminate MOST of the problems with level base and level driving surface.
also 16’s 2016 swerve, it looked like normal swerve but handled obstacles better than half the pneumatic westcoasts from that year
1 Like
The suspension lives on to this day.
2 Likes
2018 did not require a huge amount of mobility if your intake was flexible enough to take cubes from different orientations. Tank drives were hugely successful that year, and although it often is considered to be a flat field, the platform made that largely untrue. My team had a drivetrain with omni wheels on the corners and broke our omni wheels by hitting the corners of our practice field ramp. This did not occur at competition, but a drivetrain that could cross that ramp without difficulty was important. Between houston and detroit, the winning alliances only had one non-tank drivetrain out of the eight robots (2767, who used swerve).
I would argue 2022 was more dominated by omni directional drivetrains; cycles took a far more circular pattern. Even so, a tank drive robot, even without a turret, could perform at a very high level of play (better if the shooter was on the opposite side of the intake). Take a look at 6328 specifically.
When you go for a drivetrain like x-drive, you are limiting yourself to offensive roles. Any drivetrain that vectors part of the force in a direction other than the direction of motion will not be as fast or as pushy as a properly built west coast or KOP drivetrain. A very well designed x-drive can be made to go fast enough, but it is likely beyond the expertise of a second-year team.
This isn’t fully accurate. My team collected data from our scouts in real time, so we knew what our opponents had gotten in their last match. A high-scoring robot with a new number would still get defense.
Another thing to note: by selecting a unique drivetrain like x-drive you limit your pickability as a defense bot. In alliance selection, a robot who does not score well will not be picked if they don’t have a pushy, fast drivetrain. KOP is often preferred on the picklist for new teams as every older team knows how to help with one and they are very battle-tested and durable. West coast drive is a little less foolproof, and takes more work to develop, but does have potential for more pushing force and faster travel. We keep data on the reliability of a robot (if they became unable to drive) and use that as a factor in our scouting decisions for defense teams.
Tank drivetrains are also far better at autonomous driving than omni/mecanum based drives, partially due to the difficulty of collecting accurate odometry (1986’s robot in 2017 had a solution for this, although it does add complexity) and the lack of documentation for unique drivetrains.
Tank is also cheaper; mecanum and omni wheels can get quite expensive, and having four gearboxes instead of 2 will eat up quite a bit of funds as well. The KOP drive is $600 from Andymark (that’s what the voucher for opting out is worth), and has very minimal machining requirements.
TLDR: Go with KOP. It’s more durable, more versatile, faster, and pushier. It’s easier to build and will allow you more time to build a good scoring mechanism, program and get enough driver practice. It also makes it easier to build a strong auto routine.
6 Likes