StangPS Presentation Released
 

For those who don't know, WildStang (111) used a navigation system dubbed WildStang Positioning System (StangPS) during our autonomous mode.

There have been several requests for the presentation that we had in our pit area at both the Midwest regional and the Championship. I am pleased to announce that we have uploaded it to our website. There are 3 resolutions of Flash and 2 of AVI. The AVIs aren't as nice as I was hoping they'd be, so I suggest using the Flash versions.

See the link labeled "About StangPS" on the left sidebar of http://www.wildstang.org

We are planning on releasing a white paper at some point in time detailing the system, but the presentation will describe the basics.

Dave

You guys wouldn't want to post a picture of your drive train would you (or explain what you used)? I would like to possibly come up with a design for a 4 wheel steering system like your robot featured for future years.

Telly

You can find pictures and sketches in our Inventor submission.

http://www2.wildstang.org/2003/inventor/main.html

One of the other guys on the team was planning to write something up, so I'll let him give the explanation.

I must say i am very impressed. Our team has tried doing the crab/4 wheel steering in the past and we know how hard it can be to get that...you guys are by far the masters of that style of stearing. And now to add this positioning system is amazing. Very very nice!

Will we be seeing any documentation related more to the actual specifics of the custom circuit and the programming and that kind of stuff?

Patience...
We're all trying to catch up on real work after neglecting it since January. I think each of us plans to write up a whitepaper on the design of the part of the system that we worked on. We're also talking about holding some kind of presentation/class on software (including crab), autonomous mode, and custom circuits. If there's enough interest we'll try to make it a reality, so let us know if it's something you'd like to see.

Telly: We're in the process of putting all our pictures on the web; hopefully there will be some good ones of the drive system. I'm also working on a writeup for the crab software control.

Mike

Thanks alot, anything would be appreciated.

Thanks Mike. You know I'd be interested in anything you guys would be willing to share about your team's robot and its systems from this year ;). I was sad that I had to wait until the finals to see the 'bot in action, since I was on Curie the rest of the time. Just like last year and the year before, I'm in awe of your team's robot. Keep up the great work Wildstang!

Very impressive indeed...I especially liked the software for recording autonomous modes. Earlier in the season I made a similar program that captured joystick input so autonomous mode would duplicate pre-recorded driver movements, but ultimately we stuck with a simple (and reliable) gyro-based program.

BUT:

I have to ask this...did the team's students ever actually have a part in any of this? No offense is meant, I'm just curious...this is the sort of system that NASA would be proud of.

Hmmm, I think you're misinterpreting something. We don't record driver's movements. We use a software program that shows a map of the field and just select points on the map. Those points are downloaded into the robot controller and our autonomous software drives from one point to the next. This allows us to create new autonomous programs on the fly in minutes, without any practice time with the robot or a field.
Yes. Our goal is always to have the students do as much as possible. This year that meant the students worked on the wiring of the custom circuit and writing the robot controller software (the RC software is by far the most complicated part of the system, BTW). A lot of the things that we've done over the last 2 seasons with the custom circuit required that the engineers learn a lot as well. As the engineers get more comfortable we will be able to better teach the students and hopefully work toward a goal of having the hardware, custom circuit software, and robot controller software completely done by the students.

to Wildstang

WOW
that is a very impressive system. Once again I am amazed at the kind of innovations that come about in FIRST.

A white paper on the wiring/ how wilddraw works would be cool

once agan, wow, and congrats on being national champions

by the way, what do you use to measure the distance traveled, and what happens if the robot isnt moving while the wheels are spinning, wouldnt that throw off the system?

No, I understand completely what your software did - I mention mine only because it (in a way) pre-plotted desired positions on a coordinate grid.

I think this has been asked before but I really don't remember, did wildstang have and problem with slipping between the carpet and the tred on the wheels? I would guess that you guys didn't, and if you did how did you account of this in your distance tracking?

Each "tick" of a rotation sensor would correspond with a distance travelled equal to the circumference of the wheel, as I understand it.

As for wheels slipping, you may be right but a measly plastic bin won't skid them any measurable amount.

Their system is the best one for this year's game.

As was mentioned, we use a wheel with 2 light sensors mounted on it (2 sensors allow you to determine both distance and direction of rotation). As for the wheels spinning without the robot moving, this is a problem to consider however in practice it doesn't happen unless we change direction on the ramp grid (causing a momentary wheel slip) or if we run into something completely immovable (like the field border).

The rotation sensor actually is measuring a distance much less than the circumference of the wheel. Our resolution is under two inches per tick and with the tread material we use slippage is kept to a minimum. All in all the accumulated error due to slipping in a 15 second period doesn't amount to very much. If the auto mode was longer than 15 seconds we would need to use something else, perhaps a fifth wheel. A bigger problem is that with a rigid frame and no suspension, all the wheels do not stay on the ground when hitting the ramp at an angle. Our sensors are only on one wheel at the moment and perhaps sensors at every wheel averaged in software would make things more accurate. Of course that would add weight and we are already at 130.