Team 467 Driving while holding the ball

[Tsom467]

The ball is fairly well locked in place too...

http://www.youtube.com/watch?v=a30z-o4CAqc

[Jeff Rodriguez]

That robot better be named 'Scorpion'.

[cloud_254]

:D i seriously agree it should be named scorpion

[octothorpe]

Cool design, I like the rounded-triangle base. How do you plan on outfitting it with bumpers?

[ggoldman]

Firstly...very nice design.

Secondly...just like when a fighter gets a ring fighting name....your robot should be called scorpion...because thats what the PEOPLE will call it :)

On to serious business...

I noticed that your robot has a point when you accelerate when the high COG (center of gravity) causes you to pop back. If you accelerate too quickly you might run the risk of falling over.

One way to combat this is to prevent very hard accelerations with you motors. You can do this by implementing a "ramping" algorithm into your motor code.

Basically, you program a shape to a ramp (either linear in the form of

motor speed = joystick_value *(x)

where x changes from zero percent to 100 percent over a set time period

You can also do sinusoidal ramps

motor speed = joystick_value * cos(theta)

where theta goes from 90 to 0. This will give you more of a ramping curve.


Hope this helps you some!

[BenB]

You could also do a power ramp (e^x) or a quadratic (x^2) or a cubic (x^3). The posibilities are really endless...

[Guy Davidson]

You could also do a power ramp (e^x) or a quadratic (x^2) or a cubic (x^3). The posibilities are really endless...

Ahh, but they all return roughly the same result :P The easiest way to implement an exponential curve, I believe, is by using the Taylor (or Mclaurin, in this case) expansion. In the case of e^x, the expansion is 1+x/!+(x^2)/2!+(x^3)/3!... I would not be surprised if calculators do this same thing to compute powers.

[Tsom467]

Re: Scorpion... LOL! That's what one of our team members said it looked like too! I have a feeling that is what it will be called.

Re: COG... while it did "pop" back a bit, it's actually quite stable.. there's plenty of weight in the bottom (the last weigh-in was 91 lbs), but there's still more to add, including the "skin", some more wiring, and we've already added more strength to the drive sections (all custom built, btw).

Re: ramping... we do have a ramping algorithm, that we've used successfully for the past couple of years, but I'll let our programmers comment on that. You can see last years (Rack & Roll) robot in the background too.

[Tsom467]

Here's another one:

http://www.youtube.com/watch?v=rtEbJ1zsqgc

These are all in low gear, by the way.

[Jimmy Nichols]

Have you drive it on carpet yet?

[RMS11]

looks good but a video of the max hight would be awesome:cool: .

[Tsom467]

I'll get one soon, but be assured that it reaches well above 6 1/2 feet :)

[Leav]

who is the guy with the cool beard?
how does he work the lathe?? :ahh:

[Tsom467]

Very carefully :D

[Caio]

that is an amazingly slick looking robot...

i'll have to third, or fourth, the fact it should be called scorpion :)

[Jaybee1405]

Sick!!!! Are you guys coming back to FLR? Maybe we'll be pit-neighbors again lol

[rachal]

thats a SWEET chassis. are there clear pics of it anywhere? :p

[psy_wombats]

Sick!!!! Are you guys coming back to FLR? Maybe we'll be pit-neighbors again lol

We'll be at Hartford and Philadelphia this year. Sorry.

Here's a clearer picture:

http://tsom.homelinux.com/gallery2/d/13241-2/IMG_3245.JPG

[Tsom467]

Here's another video, this time picking up the ball:

http://www.youtube.com/watch?v=w8VrYpOvwFQ

(ultrasonic sensors trigger the grabbers, though we're not 100% sure we'll be doing that - but the sensors do activate a series of LED's that show distance to the ball)

And here's a close-up of the chassis:

http://tsom.homelinux.com/gallery2/d/13278-5/IMG_3260.JPG