Subject:
|
Re: Wireless Robots
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Mon, 22 Nov 1999 08:56:12 GMT
|
Original-From:
|
Christian Jacobsen <xtian@SAYNOTOSPAMnet-collect.com>
|
Viewed:
|
586 times
|
| |
|
|
Hi Pete,
I have been reading about AI concepts a lot lately (and my partner here at
work has his doctorate in AI), so I have some thoughts about your comments
below:
> 1200 baud, it takes too long to get data to the PC, and get the
> response back to the robot. e.g the sensor detects a wall, but by the • time you
> get that info to the PC, and get back a command to turn, you have already • hit
> the wall.
[snip]
> I was trying to do all the processing in the PC. The technique might work • if
> you found a way to split the work.
The current path in robotics seems to be to put local intelligence (like
walking and obstacle avoidance) into the legs of the robot, and then the
higher goals into a central "brain".
Applying this principle to LEGO, my thought was that the basic RCX
programming would take care of moving the robot around, just like a normal
MindStorms robot (or like the Tank described in the O'reilly book). Then
the processing on the PC could deal with the "higher level" things like
tracking a person/other robot, mapping the terrain, etc.
These functions are not so time critical, and therefore it makes sense to
move them off to the PC, and let the RCX deal only with the mobility of the
device.
That's the direction I am going to head with my MindStorms (when I get
them...).
- Christian
|
|
Message is in Reply To:
 | | Re: Wireless Robots
|
| Matt, I have built a lego robot controlled by a PC, using an IR link to communicate commands. It used the same IR technology as Mindstorms, but had an omni-directional IR link on the robot. The problem is link speed. For reliable communication, you (...) (25 years ago, 21-Nov-99, to lugnet.robotics)
|
12 Messages in This Thread:
 
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
