Re: WCD vs. Swerve
 

What you guys have done is exactly what I REALLY wish for our Team to try and pursue! Most-likely not this year, and perhaps not even in 2013...

I am going to approach this a bit differently... where the driver presentation is the very first thing we are going to get nailed down... if it cannot feel exactly like I want it then we shouldn't bother making it. In short I submit an idea that most people will not believe... here it is... I believe you can make it feel like a tank arcade drive (or even tank steering), with strafe.

There is one other idea I want to throw out... and that is the fear of failure hurts new innovation. I am prepared for the risk of failure with this driver presentation and I am not afraid... I know it is something I MUST pursue... it is like a calling that I cannot ignore.

Re: WCD vs. Swerve
 

I'm curious why you think most people will not believe this.

Re: WCD vs. Swerve
 

Same here.

I have seen swerve perform just like a tank drive. It's kind of pointless in my mind. But definitely possible.

Re: WCD vs. Swerve
 

There is much more to the idea that I cannot seem to put into words... for now I'll step down and get to work on it, and perhaps post a simulation demonstation sometime this summer. I really want to make sure this can work before I say too much. I guess for now let's just say I am talking to myself about believing this can be done. :)

Re: WCD vs. Swerve
 

Ok I got to thinking... here is a game I've been working on since 2007... it can hopefully convey the idea I am talking about.

http://www.termstech.com/files/TheFr...CarpetRide.wmv

In short... we spend many hours to make sure the Joystick to control the ship is as easy to use as the mouse.

Re: WCD vs. Swerve
 

It would be good practice to try.

An inability to communicate your ideas effectively is a handicap you would be well served to strive to overcome.

Re: WCD vs. Swerve
 

Agreed... but not today. I need some time to prepare.

Re: WCD vs. Swerve
 

Driver presentation. May be some who have actually driven a swerve will comment. We have a 4 wheel drive 4 wheel steering bot. We have used a x-box and 2 kop joy sticks. Also, we tried a joystick with twist but, the drivers hatted the twist for chassis orientation. So how do you control the 4 degrees of freedom required for swerve driving. X, Y, Chassis orientation, and velocity.
We have always used the left joy stick for x-y and extrapolate velocity from it. The right joy stick x mixes in chassis rotation. This is what the programers and drivers ended up with. I feel that after watching our driving the last 2 years there is a major problem with this choice. Our drivers can make the bot dance on our practice field with no pressure. Under pressure at a comp I see the driving deteriorate. I believe their left hand or thumb coordination is being overloaded. What have other teams used. I believe the extrapolated velocity is the problem. For a short time in the 2010 off season we had the X - Y on the left X-box controller joy stick. Velocity on the right joystick x and chassis orientation on the analog triggers. I liked it. The programmer graduated and the code disappeared. We went back to the above described method. So what is the best driver presentation. Arguments welcomed.

Re: WCD vs. Swerve
 

I think the most important design decision is to fundementally separate control and implementation in such a way where it becomes trivial to switch things around for trial and error. Also make it where there is a desired movement lead what is physically possible. I have solved these problems already where I fundementally work with a 2D vec of desired velocity and a float for heading. I have a class that figures out how to make that work. With this approach I can achieve the feel (e.g. elastic bell curve on rotation) of what the ships do in the game demo link I sent.

Me personally I think one arcade drive joystick just like tank with some strafe buttons elsewhere... but I want to customize to what the driver wishes.

Re: WCD vs. Swerve
 

Ok I will take a stab at this today... with a question. What makes a swerve drive so hard to drive vs. what makes a WCD easy to drive (both tank steering and arcade configurations)?

Re: WCD vs. Swerve
 

I will assume that, in this context, by "tank steering" a swerve you are referring to the driver interface (e.g. using Y axis of left and right joysticks) and not to the inverse kinematics used (i.e. not skid-steer).

Instead I assume you mean something like this:

FWD = (Y_L+Y_R)/2

RCW = (Y_L-Y_R)/2

STR = 0

... where Y_L and Y_R are the (inverted) joystick commands, and FWD, RCW, and STR are as defined here.

In that case, I will answer your question with a question: is a swerve with that driver interface "so hard to drive" ?

Re: WCD vs. Swerve
 

Thanks so much for that link! That is half the problem I was trying to figure out. The other half will be the reverse of that where given the wheel angles and speeds, what is the current x, y and rotational (i.e. angular) velocities.

For now let's ditch the tank steering 2 joysticks except to say that it can be done. I think FWD and RCW can be on one joystick where left and right perform the rotation (I believe this is called arcade drive)... Just this much is what we had this season on a WCD, and it felt intuitive (we played defense). Now add to this some strafe buttons (and not another axis). I think for me personally I'd like this because this is similar to how games like ut2004, quake etc... work. Except they use a mouse for the orientation. The strafe buttons work where they inject more strafe the longer they are held down, and then release it in the same manner. This way if the driver doesn't want to do it... it is easy to focus on the basics.

One good way to really answer this question is to create a simulation and give it to a real student driver and let him decide if it is easy or not. I *hope* to do this... next summer. ;)

Re: WCD vs. Swerve
 

But then the strafe would be digital, not analog. Not saying this is good or bad; it's just something to be considered.

Re: WCD vs. Swerve
 

That is called the Forward Kinematic Problem, and for swerve it has no kinematic solution for arbitrarily chosen values of the wheel speeds and angles. See the discussion starting at the bottom of Page7 of this paper.

See the list of driver interface suggestions on Page 1 of this paper.

Re: WCD vs. Swerve
 

That all depends on how you look at it... It is possible to swap out a button control on functionality that uses an analog intensity parameter. The idea is that the button starts digital, but then transforms to analog given the amount of time it is held down. I hope that makes sense.... it is like taking the joystick and moving it from zero to full intensity in a given amount of time.

The most important point of this is that if designed right it is easy to swap buttons with axis controls with minimal code change... or overhead. This is why I love c++. :)