Our team has been thinking about switching to swerve drive for a couple weeks, and the vote has been 50/50
the pros of us switching to swerve drive would be we could maneuver the field a lot quicker, and win more pushing matches as opposed to using a west coast/tank drive system
the cons of us using swerve drive is that our team is based in a really small town, so we will have to do a ton of fundraising, and we have decided that if we do in fact switch to swerve, we would put it on one of our drive-base robots so our drive team can learn how to maneuver the drivetrain.
If it helps, our team has been a defensive robot since the team started in 2016, so swerve drive would probably switch us to more of an offensive/support team instead of almost strictly defense. Also, we could pay for some of it from this years worlds funds, since our team didn’t qualify.
Well the offseason is a good time to develop and test new ideas. One thing to consider, since you mention you’ve been a defensive bot for a while, is working on prototyping some of the common FIRST designs so you can potentially have a more offensive bot when the season comes around.
Intakes, manipulators, lifts, and shooters can all be prototyped in the offseason as a way of adding a concept to your design arsenal for when the new challenge is revealed.
So, I really do think swerve is the future.
But should it be your present?
Looking at 5920’s history on TBA (which oscillates between second picks and blue-bumper captains reaching quarters at the district event level), I’d ask yourself what else you could develop or improve in your process instead of developing swerve. When AndyMark drops the AMSwervyBoi4U in the kit*, those other developments transfer. If you get too locked into your own homebrew swerve development, you missed those other drivetrain-independent development improvements and the masses got a big lift.
Maybe you develop the bandwidth to attack multiple things at once. If not, you’ll have to decide which meets your goals better. And maybe that is swerve, but you’ll have explored it from a better angle (since the relative value of swerve can fluctuate with game design).
*I have no knowledge of if this will be a real thing, just an example of a hypothetical day that we get swerve in the kit instead of 6WD.
Switching to swerve drive requires a lot of changes and adjustments. You can’t really just add swerve modules to an existing tank-style drive base, you would need to design a drive base around the modules you choose to use. There is also the issue of effectively programming your drive base; there are lots of other threads here on CD about the challenges with that. And as you pointed out, there is the cost of the modules themselves, plus the cost of motors (like the Neo or Falcon 500), if you want to get the most performance out of your modules.
The difference between swerve and not-swerve in terms of performance isn’t quite the magic that it sometimes looks like. There are any number of event-winning, offensively-focused robots this year who ran not-swerve. My favorites are 973, 4481, 987, 179, 3374, 7641, and my own team’s, but look at most events and you’ll see top performing robots without swerve. There are also numerous examples of robots with swerve that performed poorly, sometimes because of driver skill, others because of mechanical or programming problems.
In the end there are advantages to switching over, but they come with pitfalls and traps. Make your choice carefully, and with full consideration of your team’s abilities and limitations. See if there are other areas of growth that you can focus on without the major investment in swerve, and consider working on those first.
Good luck to y’all!
When this happens please call it this.
There are 3 reasons to use a specific Drivetrain:
- It’s the one you know how to build the best
- It’s the one you know how to program the best
- It’s the one you know how to drive the best
These 3 reasons outweigh all other.
My suggestion: If you are seriously interested in doing Swerve: Go for it! Take the offseason and build one. However, if you can’t reliable say by build season that you know swerve in and out better than a KOP chassis, then don’t switch to it yet.
You don’t have to make hard decisions about your drivetrain for next year yet. Put yourself in a position that you could theoretically build it, then make the decisions based on hard, actionable facts this December.
I hear that your goal is to become an offensive robot, and that you are considering swerve as a pathway to it.
Your level of achievement is blue captain (top 10), “primarily defensively”.
To be a blue captain off defense tells me you are plenty good at field movement, and probably even get drive practice before your events.
Coming back to “we aren’t an offensive robot” and the desire to “solve” this by adding engineered features… Have you ever built a turret you’re happy with?
What is driving the need to skip building 1 turret, and go straight to building and fielding four turrets at once? Is it worth simultaneously throwing out your most successful subsystem to date, or are there other areas to take steps forward in first?
Stepping back one level, this is how to succeed on offense in California:
- Driving reliability (drive every match)
- Driving practice
- Driving practice
- Superstructure reliability
- Superstructure features
- Driving features (sideways)
(Simultaneously, I am guilty of buying two thrifty swerve modules to play with adding driving features, and those two modules got built and now have been collecting dust for the past five months while we had our first ever finalists run this year on tank drive. YMMV.)
I am a fan of teams who are self-aware. But I think @s-neff hit the nail on the head here. If you do indeed have a goal of becoming an offensive powerhouse, then you have many other places to optimize in your strategy and design.
Swerves can play some killer defense too, and probably should if their mechanisms can’t out score what the number of points they can take away from the opponent alliance.
You might be interest to look into understand why you’ve been defense in the past. A good defense robot often doesn’t actually have issues with their drivetrain, as that’s their primary means of being an effective alliance partner.
This year is a huge outlier in which swerve is a huge and probably the single biggest advantage you could have (assuming the rest of the robot doesn’t suck). In most previous years, a well-driven swerve would be at best marginally better than a well driven 6wd and that advantage would only come in a few places (auto in 2020, avoiding defense in 2019…) I wouldn’t expect future games to be this swerve friendly, especially if they make a stronghold 2.0.
If you feel confident in building solid mechanisms for both shooting and pick and place games, (shooter, elevator, intake, indexer, climb, something new/weird the GDC throws at us…) then I would go ahead. Otherwise, make sure your team feels comfortable designing the aforementioned mechansims.
Generally the point of swerve drive is to not get into pushing matches, if there is a defender, spinning and swerving around the defender is much more effective than pushing through them.
Strongly disagree with building a turret instead of swerve. Designing a turret is much harder than implementing SDS modules for significantly less gain. There are tons of fine details that need to be done correctly for a turret to be equal to or better than not having one (stiffness, wire management, robot packaging, software tuning, weight…), and if any one of them is done wrong, you would be much better without a turret. Most of the turret bots I’ve seen this year would be better without a turret or better with a swerve instead of the turret, and its usually only the top level teams who have the experience to design a turret who actually do better with a turret than without one.
Mason, im going to be a little harsh here, but having lived and competed in Eastern Washington for the last 9 years, I want to see the teams in the region improve.
I do not think you ready for swerve, and I wouldnt bother putting the time and money into it when you can gain much bigger insight into doing what Peyton said and working on past common mechanisims (shooters, elevators, climbers, etc) instead of trying to explore the “hip new” stuff like swerves and turrets. Both require some serious programming and mechanical skill, and even if WPILib and COTS parts have made the barrier to entry much lower, they are not 0 and do require some good FRC knowledge to package and make full use of them.
What id highly recommend is taking the offseason to pretend you are back in a year such as 2018 and 2019 (you likely have game pieces from those years still) and just do a full on mock build season as the budget allows, or focus on a specific part of the bot and try to build that. This will be a much bigger help in training students on workflows and how to build stuff for FRC specific applications.
Id also recommend looking at 3636, 7034, and 7461 from this season. You do not need swerve to be good, even in PNW where it seems like every other bot has a MK4 slapped onto them. 3636 and 7461 are part of the open alliance, which means they share their whole build season for everyone to look at and critique/make use of. I highly recommend reading their threads from this season.
I didn’t say design, they can throw a Greyt clone on first 
The point of my advice is to go superstructure first, and take on the level of complexity you’re warning them away from, while still having a fallback plan to driving and defensive smarts at comps.
I’m skeptical that putting on swerve modules is as easy as you say, but maybe a better analogy is a bolted down shooter that works at multiple ranges and locations, then?
I’m going to take a different stance than most of this thread.
Do the benefits of constructing a swerve in the off-season bring value to your team? Note this doesn’t have to be on field value - perhaps you have a contingent of students who are enthusiastic about it. Perhaps it is a good test of fabrication or design processes.
If you want to and stakeholders are engaged - do it.
Perhaps gamify it in the sense of “well the first step is doing a comparative analysis of swerve and understanding how it would competitively benefit us” and then you have the analysis in your back pocket and students have been exposed to a range of how teams use it. But also their construction techniques and maybe even processes.
There’s more to frc than banners - sometimes doing something more fun is better.
Swerve does not make you a better offensive robot if your drivetrain isn’t the limiting factor in your scoring ability. While this year swerve has been a big deal, in past years it hasn’t. In fact, from 2011-2019 (except 2017), no captain of an Einstein winning alliance (in houston champs for years that applies) has been running anything other than a differential drive. (I’m fairly sure it goes back further than 2011, but image resolution drops off pretty fast and I can’t tell very well. Also, I may be a bit biased about this as my team ranked 1st at our regional with no turret and a kitbot chassis.)
Details as to who these teams were if you would like to verify
In 2019, 973 was the captain of the alliance that won worlds… with tank drive. 254 ran tank and won in 2018. (They also won with tank in 2017, but not as captains). 330 won with tank in 2016. 118 won with a tank in 2015. 254 won with tank in 2014. 1241 won with tank in 2013. 180 won with tank in 2012. 254 won with tank in 2011.
What set these teams apart? What goes on top of the drivetrain, and how you use it. The difference in swerve is minimal; many of the top teams, even this year, aren’t using it. Swerve is better, but it isn’t better enough that it will make a difference. Developing a high-speed elevator or highly accurate shooter mechanism, getting a turret, or testing a different main design strategy for your robots is probably going to be cheaper and more helpful for your team.
Still have to get many little details correct including software and wire management and packaging that can’t be solved with COTs. Getting these correct is quite complicated which is why few teams actually improve with a turret.
+1 to the multiple range shooter however.
Our team may be an outlier but we didn’t have that much trouble putting on swerve modules. We did a little (not enough) experimentation with swerve in the offseason but somehow didn’t struggle much with it in season. SDSlib does deal with most of the swerve software that historically was a very challenging part of swerve. The only struggle I remember was getting pathfollowing working with swerve but that all that did was prevent us from having a 5 ball week 1.
2767 was the winning captain of St louis and festival of champions in 2017, however, they have years and years development on thier custom swerve. And I don’t think swerve offered that much benefit to them that year either.
going to swerve could make a you an even more defensive robot, at 10,000 lakes my team got picked by the 2nd seed alliance for our defending capabilities with swerve when the rest of our robot was unreliable the rest of the event, in the end we made it to finals and were beaten by 1st seed, swerve is definitely expensive though, for our bot we used falcon 500 with SDS MK 4 modules, it was really expensive for just our drivetrain. I don’t know the exact cost but it was at-least >3000, but if you can get the funding it’s definitely a great drivetrain.
As a counter argument to this…I tend to favor KOP chassis on our 2nd pick list. Especially for defense.
If we pick a robot that is running the KOP chassis, we know exactly what we are getting. If they have a swerve, even a COTS swerve, that adds enough failure points that I really get concerned. Twice as many motors. Extra encoders. etc. It’s hard to be sure that robot wont fail in elims.
I’m not saying I wouldn’t pick a swerve defense bot, but in my eyes, it puts them a bit lower on the list.
$1240 per set of modules, $240 for the encoders ($320 if you don’t want to solder), $1200 for the Falcon motors is $2680 to get one set of modules, shipping not included.
Then there’s spare parts. You probably want a whole extra set for your programmers, plus several hundred in spare parts or spare modules.
If you really want to make a “best” drivetrain for winning pushing matches, wheel upgrades for the am14u5 to get 8 6" performance wheels with three falcons per chassis side, on a shifting gearbox, is $2188 plus shipping.
But you don’t need that to perform well; a standard KOP drivetrain is good enough for strong offense, defense, or even trajectory planned auto (with encoders and gyros of course). Your drivetrain is good as is. What you put on top of it will change your robot’s performance far more.
If you’ve always played defense as a tank bot, swerve isn’t the silver bullet to suddenly become an offensive powerhouse. I would get comfortable with building offensive tank bots before deciding it’s time to go swerve. Mikal’s advice is a great place to start. If you’re comfortable with the AM14U chassis but budget is still a concern, moving to a custom West Coast Drive chassis would be the next step, and i would stick with it until you’re sure that you can build robots that will keep up with the competition. A tank robot playing good offense will always stand out more than a defense only swerve bot.
This. 2191 was known within the FMA district as a strong swerve defense bot in 2019 and 2021 (though unlike 2019, we had some offensive prowess this year, scoring double digit cargo on occasion). While that has some value, and we don’t regret going swerve at all, just be sure you know what you’re getting into. If you’ve been primarily defense, it’s probably because your robot intaking and/or scoring mechanisms weren’t consistently effective. Swerve can help a little with those things, but better design will help a lot more. Take it from someone who’s been running swerve since 2018, you can absolutely still get funneled into a defensive role with swerve, and that’s not really somewhere you want to be.
Of course, drivetrains are the offseason project most likely to be useful the following year. But you still might find it more useful to compare the really efficient teams’ intakes to your own, understand why they’re so effective, and try to replicate that in the offseason instead. Even if the exact design can’t carry over, the lessons learned about what makes the intake work are still valuable.