So I have come up with an idea for a base and as far as I can tell it has never been done before. Now for the record I’m not a big builder or designer (I stick to driving and assist with some programming and outreach) but I got to thinking about a problem my team is facing. We are having an issue with swerve drive and I thought of the idea of having a triangle-ish base. It’s a triangle but the angles are flat (so technically this is a hexagon). My idea was to put omni-wheels on those flat edges attached to Falcon 500’s and then a single swerve module in the center. In thought, this will give us the ability to still strafe around and have the movement of swerve but without the complexity of the task. Wondering if anyone has any insight or thoughts on the idea.
Apologies for possible rambling. This is my first post. Feel free to ask any questions and I will attempt to answer them to the best of my ability
While I’m not sure your exact idea has been done before, it has been done with just the omnis in the corners. It is called “Kiwi Drive” and it saw some success in 2015 but is generally not very viable. As a general piece of advice, if you can’t do swerve, do tank.
In general my team does tank, in fact we will probably will this year as well, this was a side project I’m working on to see if it can work and spend time to try to perfect it. I appreciate your insight and will look into “Kiwi Drive” and think further.
The “fit inside a goal” was a reference to team 1501’s 2010 robot, by the way. They got stuck in one of the goals that year, fitting pretty well as I understand it.
This setup sounds more difficult than a traditional swerve drive for worse performance and I’d highly advise against it.
The described setup without the swerve module is often called kiwi, and provides all the omni-direction movement of swerve. It is usually not done (with some notable exceptions) due to the amount of traction lost compared to swerve (omni wheels don’t resist pushing in the lateral direction).
Adding a swerve module in the middle is unlikely to help, as you’re adding a single driven wheel with traction that’llLimelight, an integrated vision coprocessor only have 1/4 of the robot’s weight. Additionally you’d essentially have to program a swerve drive and a kiwi drive to work in tandem.
I understand. Thank you for the help. I’ll look into it anyway (because I hate the idea of giving up) But your insight is not going to fall on deaf ears. I will use your advice and adapt.
Your idea can only behave the way you intend if all the wheels are in contact the floor.
Please consider what happens when the robot changes direction rapidly. Please also consider the effects of a floor that is not sufficiently flat. In some games, there are portions of the field elements that attach to the carpet, guaranteeing that the floor will not be flat in that area.
It would probably require some sort of suspension system on the omni-wheels. Getting the suspension right will take a lot of work and you would be on your own trying to figure it out. You would be better off learning how to do more fundraising and buy a set of 4 swerve modules.
I’d encourage you to work with your programming team to better understand and describe the issue. For example “when we command the robot to go right, the front right and back left wheels do not point in the direction of travel”, with some relevant snippets of code, etc may get you a list of things to check.
There is always merit to trying new mechanical implementations from an educational standpoint, but the reason why the community has gravitated to the “standard” swerve design is that when implemented properly it works well. It sounds like you’re just having an issue with the implementation and you should focus on fixing that issue, versus trying to solve the problem with a new design.
I understand… My thought process was that every drive train system was new and unique at one point. I figured if someone else could create something new so could I. Appreciate your guidance and will advise them of this shortly
