PWNAGE did shifting swerve in 2015
1 Like
There are a few teams that have done shifting swerve 2767 did one (2013 I think) and there are a few teams that still use them that are skipping my mind.
There are even some teams out there with CVT swerves. (maybe a few people can link in a few examples of the top of their heads)
1 Like
1640 is probably most notable.
2 Likes
Here are my two outlandish ideas:
1: swerve & tank drive. Having some form of drive train that can switch from tank to swerve an visa versa, creating the ultimate power and speed combo.
2: (this may be unrealistic) Adding an additional motor to swerve movement. With my minimal research the torque of the movement of swerve is doubled with a second motor while keeping the speed the same. Now, this may seem impossible or overbearing with this idea relying on 16 motors, but there are three ways I can think of doing this; One, is creating a triangle swerve with 9 total motors (3 modules, 3 motors per module, 2 for driving, 1 for rotation) but I don’t know how that may affect the stability/programing. Two, adding a second motor to only two of the modules (again wont affect speed but may create issues with programing). Or three some form of PDP Extension that would be developed by FIRST in the future. This is idea is outlandish but perhaps possible.
1 Like
It would either have to get a whole lot cheaper to produce or FIRST would need to find a sponsor willing to cover the increased machining and material costs. I guess some combination of suppliers could also partner up to distribute costs among themselves.
The controls would also have to get to a point where they can be just as (not almost as, not just as with enough experience) simple as a tank drive - which I doubt could happen in the near future due to swerve having the complexity of the turning motors while tank drives are as simple as setting the power on two motor controllers.
In 2021 many teams did 3 motor swerve modules. 1678, 1619, 148, etc… (a lot have videos up on YouTube). I personally did a robotics research class where I built one and posted progress on here.
1 Like
Yes, but what about a 4 motor swerve module!?
1 Like
On It. Lol
4 Likes
RIP batteries.
2 Likes
Diffy swerve with suspension for terrain or tri motor Diffy swerve but that pushes the batteries to the extreme limits
2 Likes
I personally don’t see this happening at all. A crucial attribute of the KOP chassis is that it allows even a rookie team with very little mechanical/programming knowledge to have a fully functioning drivetrain for competition. Unless there can be a way to drastically simplify the mechanical/programmatic complexity of a swerve drive, it won’t make sense to have the KOP chassis be exclusively swerve.
5 Likes
With diffy swerve would you really need 3 motors?
I do think diffy swerve with a suspension would be top tier though. Now if you add in easily changeable wheel sizes and some traction control software I am not sure there would be a better drivetrain option.
2 Likes
3 motors isn’t required it’s more of just the limit you can have without murdering a battery, but suspensions and easily swappable wheel sizes would make it interchangeable year to year for any game.
three motor diffy is kinda sorta mechanically impossible tho, so like
1 Like
If you’re doing both, you might want to consider octocanum. It gives you the omnidirectional drive, the beastly tank pushing machine, and you can do it with only four motors (and more if you hate your batteries). By using different sized wheels and/or different sized sprockets, you can integrate shifting gears into the action of shifting drive modes as well, and your code is just the libraries for differential drive and for mecanum drive on a button toggle.
I realize that it’s probably unpopular to say anything positive about mecanum these days, but four motors and either two or (more likely) four pneumatic cylinders (on a single solenoid) is a much, much smaller investment in terms of money, controls complexity, and weight than a swerve cobbled (however expertly) to a tank–except, perhaps, in a situation like Stronghold, where 1533’s Swank was really a fantastic option.
Edit: please note that I did not say “octocanum > shifting swerve,” I only said “octocanum probably > swank”.
3 Likes
I think we are going to need lithium batteries to provide the current needed for any of these drivetrain concepts, 4 falcons can already brownout a robot if you don’t current limit them.
1 Like
That’s definitely true battery technology needs to evolve with the rest of the system, but you can do some very fancy software limitations of current draw and if you use a prediction model running barely ahead of the main Rio loop 5-10ms and send the data over a websocket you could create a kinematics system which only ever produces safe to apply commands by predicting the entire system draw and adjusting current limits every cycle but I’m not sure how much that would impact the CAN bus.
1 Like