Swerve for few but not for you

Technical Technical Discussion
1

Disclaimer: the title is not meant to insult anyone or their capabilities. It’s mostly a joke about an old 2017 thread, but I do think it’s fitting and somewhat relevant.

Okay SO! The other swerve thread has a lot of good discussion about the COTs ecosystem and the merits of that. But I have another aspect of swerve I want to discuss. Not about teams who are using COTs swerve to gain a competitive advantage, but about teams who aren’t. Teams who may feel forced into swerve whom are neither ready nor can afford it. Because I think this is the much bigger issue with the swerve revolution. Especially if teams try it out for the first time in a competition season.

I’ve seen far too many teams essentially driving a swerve robot like a tank drive robot, or struggling to drive at all. I’ve seen others with beautiful driving and complex autos they wouldn’t have been able to do otherwise. I have nothing personally against swerve. I’m generally very pro-COTs. But I think there are an unfortunate number of teams who are doing swerve who likely could’ve done a lot better by diverting their resources elsewhere. I guess my question is when does it become advantages to do swerve (cots or elsewise) and how do teams avoid falling into that trap?

I guess my tentative answer has to do with experience. Do swerve as an off-season project and love it? Go for it! Say after kickoff “I think this is a good swerve year”, maybe not.

Disclaimer. My team doesn’t do swerve. I don’t see us doing swerve in the near future. We have a solid WCD design which has served us well, and just got us our first blue banner this past weekend on an alliance with zero swerve robots. You don’t need swerve to be competitive, and I’m a little bit tired of this notion that you do, because I think it’s pushing some teams down the wrong path.

33 Likes
2

The rule I gave my team for swerve (that I recommend other teams use): if you can’t make it do spinning figure 8s before kickoff, you’re not using it.

51 Likes
3

Every year around this time my students ask if we do swerve, and every year I tell them if we start now, we can try to do swerve in about 2 seasons. The juniors/seniors lose interest since they won’t be around, and the underclassmen see it as being a lot of work they don’t feel prepared to tackle yet. This year we made a very robust, fast 8-wheel drive system that can push most swerve bots around, and they are seeing the value in that (as well as driving skill) over the tech. Before tackling swerve, the team needs to accomplish a highly reliable turret.

8 Likes
4

My team hasn’t done swerve before and I would love to do it, and for one reason; it looks cool.

11 Likes
5

Instead of telling them a time based restriction, have you tried a tangible goal that they can control? Like “we can do swerve during the season when you can do x, y, and z with it.” That might be more encouraging, and they may surprise you with how quickly things can get done with motivation.

Just a thought, not trying to tell you how to run your team.

23 Likes
6

I have a tangent of a question, but I do think it is relevant.

If a team buys and implements a COTs swerve system in the off season, is it legal to take it essentially as is off the practice bot and place it on the competition bot?

I look at it from two points of view and cannot easily rectify the two.

  1. It is a major (the most essential) subsystem (therefore should not be reused).

  2. With the exception of the wheels, it seems to almost entirely to consist of motors and the attached gearboxes (thefore could be reused?).

Anyway, I think this would weigh heavily in some teams’ decisions.

7

Yes you can.

  1. It is not considered a major mechanism because a single swerve module isn’t a complete drivetrain. Only when their powers are combined does it become a drivetrain. I agree that the what is a major mechanism discussion is fairly silly! I also think it would be silly to take apart / rebuild a module just to swap bots. Though with SDS that wouldn’t be hard.

  2. It is COTS. As long as you haven’t made substantial modifications to the COTS part you can reuse it.

16 Likes
8

Referring to Section 9.3 of the 2022 Manual, all the rules are marked as evergreen. The pertinent one is R302 *Custom parts, generally from this year only.

The rule is long, includes extensive Blue Boxes, and makes reference to the Major Mechanism definition in Inspection Rule I101.

My own interpretation of I101: a swerve drive system is a MAJOR MECHANISM because it is used to move the ROBOT around the field. A swerve module is not a MAJOR MECHANISM because it is a COTS item, and relatively ineffective at moving the robot around the field until it is combined with other swerve modules, fabricated items, electrical systems, software, etc.

@ChuckDickerson or @Al_Skierkiewicz , please jump in here with any relevant guidance provided to LRIs that can also be shared here. I vaguely recall that this topic is not new and may have been covered before in CD discussion, but one of you can likely point to that if it exists.

5 Likes
9

Are you saying that it can be reused assuming the modules are transferred independently or that it cannot be reused because it was once part of a major mechanism?

I agree that you cannot transfer the entire swerve drive base.

2 Likes
10

Sorry I was not clear enough. I think that “it” in your question is a swerve module, and that a swerve module can be re-used to become part of a new swerve drive system that is different in some significant way (e.g., fabricated items, electrical system, software) from the one in which the module was used in a prior year.

5 Likes
11

Last season since we weren’t really competing we decided to do a swerve drive as a project.
We were lucky that the drive we built fit perfectly into this game. We would never had used it with out the work we put in last year (we built a new one this year.)
We love it so far but we will jump right back to tank drive next year if it makes more sense for that game.;

5 Likes
12

I believe that it is important for each team to come to a consensus each year on the team goals.

A few years ago when I was mentoring an FTC team, they agreed as a team that they wanted to develop an FTC swerve drive that used 4 motors and 4 servos for steering, and that it would have full swerve capabilities (all modules able to rotate independently and continuously, etc.). That team actually stated that this priority was higher than winning. It was an engineering challenge that required a full team effort (mechanical, programming, build, etc.). Because of the FTC rules, they decided that this would be a summer effort and worked very hard all summer long through multiple iterations to develop the swerve drive. In the end this new drive train was a success. The robot’s performance was good, but ultimately it was more about the personal accomplishment of doing something that many people thought was impossible rather than winning competitions that inspired these kids to work so hard. They won the state inspire award that year, had a continuous queue of judges and other teams at their pit at Worlds and ended up sharing their design with several other teams the following year who made further refinements and improvements to the design.

It may come as a shock to some of us that a goal of trying something new and potentially failing at it may be a more worthy goal than winning a blue banner. But, the bigger question we should be asking here is what are the teams’ goals. If the team, as a team, decides that they want to stretch themselves into a design and programming challenge for the sake of the challenge knowing full well that it may end in what others may perceive as offering no competitive advantage, then they should do it.

If on the other hand, a team decides that they really want to win competitions or at least improve their performance such as becoming an alliance captain for the first time (and trust me, these are also worthy goals), then most of us agree that staying within a teams’ capabilities and not biting of more than you can chew has a higher probability of achieving that goal.

The mentors’ role in this is to help keep those goals in mind as they progress through the season and off season. If the students’ goal has placed engineering challenge above competitive performance, helping to remind them of that as they realize that their design may not be performing at blue banner levels is important. We should be helping them to see the value in pursing the engineering challenge and recognizing their achievements relative to those goals rather than trying to point out that they are not winning.

The excitement of competing and potentially winning should inspire the students to work harder and challenge themselves, but winning is not the only worthy goal and probably should not even be the primary goal.

10 Likes
13

I think of this for the games since the real boom of COTS Swerve and it heavily impacts my own interest in pursuing swerve. Swerve seems like an end game upgrade to me effectively like taking the stat% over the base stat early game.

14

I totally agree with this. I believe the suggestion of “build a swerve in the off season” makes the assumption that a team wants to maximize their points/robot potential for a single season. I think it is important to remember (and I personally forget this sometimes) that some teams have different goals.

I don’t think it would be unreasonable for a team to concentrate their efforts on a swerve drive during season, as long as they accept that this is not a way to maximize their field performance that season. Maybe the team is a bunch of sophomores/juniors who feel like they want to invest a season into a cool drivetrain, and learn the nuances of it. Maybe they just think it’s cool and want to learn.

There was 1 team at our last event that ran a “sweverybot”. I guess it wasn’t exactly and every-bot, but it was certainly inspired by it. I talked to the team for a while, and the students were VERY excited about their progress as a team. I could tell the students were inspired and excited about their robot…and i’d say that’s a good reason to do swerve.

Now if a team thinks that they need to make a swerve drive just for a competitive edge and has NOT prepared in advance, i’d encourage that team to rethink a little bit. Better stuff on TOP of the drivetrain will make a bigger difference in field performance.

5 Likes
15

My view as an LRI: Unless I have a real good reason, I approach it with a “Don’t ask” mentality. When teams tend to reuse parts, it’s all too often something incredibly integral, like the drive train or chassis. I get it - you have a perfectly good KoP drive train/chassis from last year sitting there, why not opt out, get the AndyMark voucher, and reuse it? For smaller, poor-resourced teams, this makes complete sense. When I go to inspect the robot and see last year’s inspection sticker on it, however, it raises questions (yes, this happens!).

But then, as an LRI, how do I tell that team their frame and drive train is illegal? How do I get them a new one, force them to spend the entire competition rebuilding their robot… only to end up with the exact same robot they walked in with?

I’m not saying that swerve modules are the same as the KoP drive train, but the parallels are there. R302 is an incredibly difficult rule to enforce, both because there’s a big difference between “knowing” and “suspecting”, and because of the incredibly huge impact it can have on a team for a relatively minimal difference to the event.

As a mentor, my focus is on the experience the students have. I would have a problem with my team pulling something like swerve modules off an old robot and sticking them on a new one without any other work being done. Those students didn’t have the experience of assembling the modules and really understanding how they work. I would have them tear them down and reassemble - that means I get 4 groups working on each, which means 1-3 students per module, so a lot of hands-on experience, a great chance for learning.

Now, if the team bought and built the modules in the fall, then those same students in the build season are the ones that built them in the fall. Is there a worthwhile experience for them in tearing it down and rebuilding it? Or can I justify leaving the modules intact per R302(E)(b)? In my view, the individual modules fall under that rule and can be left intact - but assembled into a frame they do not. So, at minimum, I need to create a new frame for the robot and swap them over.

32 Likes
16

We still build our own custom modules and therefore there is no question that we cannot start fabricating them until after kickoff. We have accepted that is one of the resource costs that we need to manage.

However, we have asked ourselves the question of the past few years whether it makes sense to invest in COTS modules so that we could devote more of our build season shop time to fabricating other mechanisms. When we have had those discussions, the assumption has been exactly this - that the individual modules could be fully assembled prior to kickoff and could be re-used from year to year without needing to be disassembled. Since many COTS swerve modules do not include motors or encoders, in theory, the only thing that needs to be done to a fully assembled module is to remove the motors and encoders and replace them after kickoff. The frame that they attach to is not COTS, so there is no question that the frame needs to be re-built. I guess that is how I have always read this rule by drawing the line between COTS and team-fabricated mechanisms.

However, if there was a COTS frame offered by a vendor that was suitable for the COTS swerve modules, then maybe this rule would be a little fuzzier. But, like the example of the motors, it seems to me that rules are clear that the robot needs to assembled from the individual COTS items during the build season. So even though, in this case, I would not need to rebuild the COTS frame, I would need to disassemble the other COTS components such as the swerve modules from the frame and then re-assemble the drivetrain after kickoff. If the team forgot to remove the prior year’s inspection stickers, I don’t know how @Jon_Stratis would know whether they had actually broken the prior year’s drivebase into the individual COTS assemblies and then re-assembled the drivebase after kickoff. .

1 Like
17

That’s an area where inspectors/LRI’s can very easily get a bad rap. It’s a delicate discussion with the team. To put it very badly, you’re asking questions while hoping the team doesn’t admit (through ignorance) that they broke the rule. To perhaps put it better, you’re asking questions to help the team understand the rules better so they can do a better job of following them in the future.

You need to show compassion and empathy for their process and the situation that led them to their solution, you need to enforce the rules, and you need to avoid doing so in a way that is, in the grand scheme of the event, pointless. In this sort of situation, more often than not those different aspects are in conflict.

I’ll give two, very specific, examples of times when this conversation has happened to me.

2013 example

The first is way back in 2013. For those that don’t know, prior to 2013, the robot sizing rules were a constant 28"x38" footprint (we had a sizing box the robot had to fit in and everything, and FIRST wasn’t going to build a new box every year). Then in 2013, they went a different route, with a 112" frame perimeter. Teams that built on last year’s frame, without reading the rules, were VERY obvious - their robots were larger than everyone else’s! In this particular case, the teams had a lot of work to do to get compliant, and the legality of their frame was the least of our concern (I believe I said at one point “We’re going to be tearing it apart in order to cut the size down anyways, at that point it’ll return to COTS and be legal”).

2022 example

The second example is actually from this year. I got to a local week-0 event and saw a robot that looked incredibly familiar. As in, from 10 feet away it looked just like the team’s 2021 robot (which was an iteration of their 2020 robot, just adding swerve to it), which I had looked closely at in an off-season event in the fall. I have a good relationship with the mentors on this particular Hall of Fame team, so it was easy for me to walk up and say “It looks like you guys brought the wrong robot!” That turned into a great conversation about how happy the team was with the previous year’s robot, how they realized they could iterate on it once again (building it all new) to get something even better for this year. So, a low-key way to initiate the conversation, generate a great discussion that lets the team show off their design process, and end up confident that the team doesn’t have an issue (and, knowing their mentors, including a fellow LRI, I was already pretty confident about that anyways).

3 Likes
18

My team spent the money and some of our time to prepare for the swerve revolution. We have a test platform with SDS Mk2 and own full sets of SDS MK4 modules. However we made the choice we weren’t prepared enough to program or maintain them if something went wrong.

Considering so far that even with a simpler drivetrain we have still have had other issues I’m very glad we aren’t doing swerve as well.

5 Likes
19

There are a lot of disclaimers you left out of your post, but without getting into those, I would sum this up with the FRC Golden rule of new mechanisms as a lower re-source/newer team:

If you cannot field one (swerve, in this case) in the offseason/if your team has never tried this before, you should not try and accomplish a new complex foreign design without understanding your capabilities to execute that mechanism properly during build season.

5 Likes
20

I would go as far as to say this applies to turrets as well. Seen too many non-functional turrets from otherwise good teams.

3 Likes