 |
|
| Science Forum Index » Cognitive Science Forum » Integration of AI and robot software... |
|
Page 1 of 1 |
|
| Author |
Message |
| Forthminder... |
Posted: Thu Oct 29, 2009 1:43 pm |
|
|
|
Guest
|
|
| Back to top |
|
|
|
| MarkWills... |
Posted: Thu Oct 29, 2009 6:42 pm |
|
|
|
Guest
|
On Oct 30, 12:43 am, Forthminder <menti... at (no spam) myuw.net> wrote:
[quote]Quick question here from a Forth AI practitioner:
If we have an off-the-shelf AI Mind in Win32Forth at
http://code.google.com/p/mindforth
and robotic control software in RoboForth at
http://www.strobotics.com/roboforth.htm
can the AI Mind send signals for the RoboForth
control software to execute in the real world,
or must both elements of the system (the AI Mind
and the AI Motorium) be written in the same
dialect of Forth? In other words, would I
have to rewrite Mind.Forth AI in RoboForth, or
persuade the author of RoboForth to recode the
control routines in Win32Forth or 64-bit iForth?
I ask this question because just today I found
http://tech.groups.yahoo.com/group/win32forth/message/14466
as a message from the author of RoboForth at
http://www.strobotics.com
which I am adding to my list of robot vendors:
http://www.dontronics.com
http://www.hansonrobotics.com
http://www.strobotics.com
http://www.trendrobotics.com
http://www.trossenrobotics.com
http://www.willowgarage.com/pages/robots
In closing, the theory of mind implemented in
http://www.scn.org/~mentifex/mindforth.txt
selects and executes a motor proposal stored
in motor memory not by sending a single signal
to the motor engram, but by letting pro and con
signals cause the motor release "capacitor"
to wax and wane in strength below a discharge
threshold. Only if the AI Mind thinks positively
and prolongedly about a motor option engram,
does the AI-Minded robot execute the action.
But first the AI Forthmind needs the sort of
motor output software that RoboForth provides.
TIA (thanks in advance/total information awareness)
Arthur
--http://cyborg.blogspot.com
[/quote]
I would investigate communicating between the two forths via wire,
using serial or IP protocols. That way, the AI forth does the AI, and
the RoboForth just does the control, according to instructions issued
to it by AI Forth over the wire. This also gives the option of the two
forth's running on different hardware; i.e. AI Forth running on a
desktop box, and RoboForth running on embedded hardware local to the
machinery being controlled.
Just my two pence :-)
Mark |
|
|
| Back to top |
|
|
|
| Jerry Avins... |
Posted: Thu Oct 29, 2009 11:17 pm |
|
|
|
Guest
|
Forthminder wrote:
[quote]Quick question here from a Forth AI practitioner:
If we have an off-the-shelf AI Mind in Win32Forth at
http://code.google.com/p/mindforth
and robotic control software in RoboForth at
http://www.strobotics.com/roboforth.htm
can the AI Mind send signals for the RoboForth
control software to execute in the real world,
or must both elements of the system (the AI Mind
and the AI Motorium) be written in the same
dialect of Forth? In other words, would I
have to rewrite Mind.Forth AI in RoboForth, or
persuade the author of RoboForth to recode the
control routines in Win32Forth or 64-bit iForth?
[/quote]
In a former life, I controlled a robot made and programmed in Forth by
IBM with a Z-80 running 8080 Forth. The interconnect was a serial port
because that's what IBM supplied.
You will find interesting control hardware with a Forth flavor at
http://www.newmicros.com/
Jerry
--
Engineering is the art of making what you want from things you can get.
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ |
|
|
| Back to top |
|
|
|
| zzbunker at (no spam) netscape.net... |
Posted: Sun Nov 01, 2009 3:17 pm |
|
|
|
Guest
|
On Oct 30, 12:17 am, Jerry Avins <j... at (no spam) ieee.org> wrote:
[quote]Forthminder wrote:
Quick question here from a Forth AI practitioner:
If we have an off-the-shelf AI Mind in Win32Forth at
http://code.google.com/p/mindforth
and robotic control software in RoboForth at
http://www.strobotics.com/roboforth.htm
can the AI Mind send signals for the RoboForth
control software to execute in the real world,
or must both elements of the system (the AI Mind
and the AI Motorium) be written in the same
dialect of Forth? In other words, would I
have to rewrite Mind.Forth AI in RoboForth, or
persuade the author of RoboForth to recode the
control routines in Win32Forth or 64-bit iForth?
In a former life, I controlled a robot made and programmed in Forth by
IBM with a Z-80 running 8080 Forth. The interconnect was a serial port
because that's what IBM supplied.
[/quote]
Well, that's not so surprising. IBM is the reason the people
who know how robots work invented USB, rather than Lawyers,
and invented Home Broadband, rather than AT&T, and
invented HDTV, Blue Ray, Holographics, and Self-Replicating
Machines,
rather than Chicago.
[quote]
You will find interesting control hardware with a Forth flavor athttp://www.newmicros.com/
Jerry
--
Engineering is the art of making what you want from things you can get.
ŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻŻ- Hide quoted text -
- Show quoted text -[/quote] |
|
|
| Back to top |
|
|
|
|
|
All times are GMT - 5 Hours
The time now is Mon Nov 30, 2009 5:47 am
|
|