Subject:
|
Re: Autonomous Robot
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 3 Aug 2000 23:44:29 GMT
|
Viewed:
|
851 times
|
| |
|
|
"Ian Warfield" <ipw47@hotmail.com> wrote in message
news:FyoKFA.692@lugnet.com...
> In lugnet.robotics, Mauro Vianna writes:
>
> Would you necessarily have to keep all the brains and map data confined to • the
> RCX? Taking the Visual Basic concept one step further, how about putting • the
> entire intelligence routine on the PC and just using the RCX-based rover • as a
> drone? Using Spirit.OCX based in Visual Basic, you could control the • rover
> over the IR link. Aside from the rudimentary aspects of sending movement
> commands to the RCX and querying the sensors, you would have the entire
> resources of Visual Basic at your disposal.
In fact I thought about that. I played with that already. Bert van Dam AI
site (http://home.concepts.nl/~bvandam) has some great ideas and examples on
that.
I did not consider this solution right now because of two problems:
1) Since I want the robot to stay on all the time, the PC should be also.
Even if that works, the IR tower may consume the battery if connected for a
long time. If probably would need an external power source. I hate to use
this kind of stuff with devices connected to my PC. :)
2) Unfortunatelly, my apartment has a lot of blind spots where the robot
could move and lose the IR link. I already tried that. Of course, I could
use a kind of stand alone program that detects this situation and them looks
for the IR link again.
That's a good and powerful solution that I may consider in the future. It
would be great if the RCX had a radio interface like the cybermaster.
Unfortunatelly the cybermaster can't talk to the RCX unless you use some
electronics. See Bert's site.
> I've been trying to do something similar with my computer. I've • downloaded
> only the essential movement and sensor routines to the RCX, keeping the • main
> program computer-based. This designates the computer as the "brain" and • the
> rover as the "spinal cord", if you will. The RCX is capable of reacting • to
> certain stimuli automatically, like stopping when it runs into a wall, but • the
> main processing power is reserved for the computer as the RCX waits for
> instructions on what to do next.
Another option is to have a RCX based program with some primitive functions.
These functions could be called by IR messages called by the PC. The datalog
and variables could provide feedback information.
> I imagine one could take this to the point where the computer builds an
> internal map based on data it receives from the RCX. It would display the • map
> on screen using the Visual Basic interface. If the user clicks on a spot • on
> the map (intending to send the robot there) the computer would calculate • the
> appropriate path and send commands to the RCX.
Of couse mapping it would be much easier with VB. I liked your idea of an
user interface. I didn't think about that.
> The computer would constantly
> poll the RCX to see if any new landmarks have appeared or if the robot is
> veering off the path to its destination.
In fact I want to start with a static environment first.Of course your idea
is tempting but I want to keep it simple so I can achieve my goals. Maybe a
version 2? :)
Right now I can dedicate time for Lego only in weekends. I will post any
achievements, build instructions and code about it if there are people
interested. The best place for that would be a site, but I prefer to play
with Lego than making a site. :)
Mauro Vianna
|
|
Message is in Reply To:
 | | Re: Autonomous Robot
|
| (...) Would you necessarily have to keep all the brains and map data confined to the RCX? Taking the Visual Basic concept one step further, how about putting the entire intelligence routine on the PC and just using the RCX-based rover as a drone? (...) (24 years ago, 2-Aug-00, to lugnet.robotics)
|
37 Messages in This Thread:
 
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
