DPRG
DPRG List  



[DPRG] Processing unit for mobile robot

Subject: [DPRG] Processing unit for mobile robot
From: Chuck McManis cmcmanis at mcmanis.com
Date: Sun Apr 20 00:56:31 CDT 2008

At 10:19 PM 4/19/2008, Randy M. Dumse wrote:
>So I have to wonder, if a robot with only a micro can be a
>functioning robot, but a robot with only a PC cannot be a
>functioning robot, which one really is the higher brain
>function? And which is really the coprocessor?

That is an excellent observation Randy.

Recently I have been playing with a Make Controller, this is a board 
that some folks put together and it was featured in one of the Make 
magazines. (You can see all the details at 
http://www.makingthings.com) One of the strangest things about this 
board is that its built in firmware uses something called OSC 
(described at http://opensoundcontrol.org) which is designed for 
musical instruments. I was curious because this board is similar to 
one that Facinating Electronics makes/made (see that one here: 
http://fascinatingelectronics.com/meascont.html) which is designed to 
"link your PC to the real world" (something a robot using a PC would 
definitely need!)

In both cases the sad thing is that mostly these are polled 
controllers, which is to say you send them a string (typically over a 
serial port (or an emulated serial port)) and they either activate an 
output or they respond with a reading from something like an ADC. But 
what they cannot do is notice that your robot just hit an obstacle 
and that your "brain" should give it something else to do.

So I've been musing about an appendage controller which implements 
something of a subsumption architecture which would, in my musings at 
least, allow a PC to give a general set of things to the controller 
that it should be commanding and basically some guidance of how it 
should respond to various sensor indications. If anyone is working on 
something like that send me a pointer.

As for the question of which is co-processor, that's pretty easy. I 
mean a nematode can do goal directed movement but it doesn't have a 
brain to speak of. The goal planning / analysis stuff that the PC 
does puts it in the "master" seat and the parts that move the end 
effectors and watch for obstacles are clearly "slaves" (or if that 
strikes you as offensive, sub-controllers).

--Chuck


More information about the DPRG mailing list

Copyright © 1984 - 2006 Dallas Personal Robotics Group. All rights reserved.
Website Design by NCC

For the latest robot news visit robots.net