This project has some versions, one of them with Hitechnic IR-Link and PF motors (remote control) and other two with RS485 communications (second and forth).

In the third version I used Xander´s driver in order to control the robot with Hitechnic IR-Link.

my website in english.

That thing is very impressive looking. Wow! I like how you did the drive for the balls. Kind of like a reverse-mouse x 3

I've been thinking of getting these: http://www.holonomicwheel.com/template. ... nomic.html, they look like they could be a lot of fun to play with as well.

Xander

You can see herehow is the drive for the balls. That wheels can be a good solution. In this videoyou can see another solution, very efficient.

Nice design. I did see that video before. The only thing about that robot is that it looks so fragile.

Xander

Hi,

I've recently managed to nut out the vector maths involved in Holonomic movement thanks to the path paved by Steve Myatt on the following link:
http://robocup.quantumtorque.com/de/robots/05striker/calculations.htm

This code allows you to send the robot off in any direction, in a straight line, for a set period of time, without any rotation or spin. What I'd now like to discuss/know now is:

What would be the best way to strafe around an object. As per this video.
http://www.holonomicwheel.com/template.jsp?p=my/VideoClips.html
(See the top video - The firefighting robot)

Problem: You have to factor left-right movement, forward-back movement and rotation at the same time.
Is there a basic mathematical formula utilising the distance from the object as the radius?
What are your ideas?

Thanks,
James

Well, if anyone can figure that one out, it's Ford Prefect, our resident mathematician. Will he step up to the challenge?

Stay tuned....

Sorry, James, but the first link doesn't work.

The holomonic wheels are a good way to do some movements, but the mathematical formulas aren't easy. I've them, I did a little robot with them. I defined some speed combinations and run fine, but it isn't easy to control the direction with precision.

I've some links with documents about navigation and they use in all cases differential direcction , but they are in Spanish (http://sauron.etse.urv.es/public/propos ... 333pub.pdf, http://gro.usal.es/trabajos/documentos/gradoJAVL.pdf)

Koldo

hi, Xander!
thank you for the compliments, but I'm really not a mathematician!
Maybe some day I'll have a look at this, but actually there are 4 things which are more important to me right now:
1.) the NAN and calculation exception compiler/runtime errors must have been fixed
2.) the BT network with mux features must work
3.) the RS485 network with mux features must work
4.) the motor commands for switching each single motor to float/coast or to hold/brake mode must be established.
Unless these 4 things been fixed and finished and I'll have got my AI-Navigator-Robot working, any RobotC programming will stay at a dead stop, unfortunately. Currently I'm figuring out some alternatives.

Here an example of my robot using 2 NXT and 1 Nunchuck:

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