Subject:
|
RE: legOS (-> Interdisciplinary Learning?)
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Tue, 1 Dec 1998 20:39:04 GMT
|
Original-From:
|
Tom Pierce <tomp@AVOIDSPAMcitrix.com>
|
Viewed:
|
2498 times
|
| |
|
|
<snip>
> > Can't the IR transceiver send numbers in the 1-255 range? That should
> > enable some fairly productive messaging.
>
>
>
> I have no idea (yet) if the IR transceiver can do that. But how would you
> use such a messaging capability? My (admittedly still very inchoate) idea
> was to simulate "real" life conditions by either using the temperature
> sensor someone has mentioned (predator) + some heat generating device
> (prey), or alternatively a light sensor/light to simulate smell (by making
> the predator reacting to a gradient in the light intensity, and perhaps by
> using two light sensors to provide some directionality).
> I haven't tested either of these ideas, not will I able to do it until the
> quarter is over and the finals are graded. But I am looking forward to
> some
> very interesting vacations. And if it counts as work....
[Tom Pierce] Some friends and I have made some use of the IR messaging to
simulate 'conversation', using the 1 to 255 number sequence. We used the
messaging to indicate eight distinct messages in each send:
X = the IR send number.
X <= 128 indicates that the RCX is on 'team one' (or, alternately, 'prey').
X > 128 indicates 'team two', or 'predator'. X = X - 128.
X <= 64 indicates that the RCX is mobile/in motion.
X > 64 indicates that the RCX is stationary. X = X - 64.
X <= 32 indicates that the RCX has (one or more) light sensors.
X > 32 indicates that the RCX has no light sensor. X = X - 32.
X <= 16 indicates that the RCX has (one or more) touch sensors.
X > 16 indicates that the RCX has no touch sensors. X = X - 16.
X = 8 indicates that the RCX is attached to a bot with more than one RCX.
X > 8 indicates that the RCX is solo. X = X - 8.
X <= 4 indicates that the RCX is inquiring/pinging.
X > 4 indicates that the RCX is answering/responding. X = X - 4.
This leaves us with X equal to 1 through 4. We use this last number to
distinguish bots on a team (obviously, in this example, there could be a max
of 4 bots on a team).
So, to sum up this overblown and excessive data transfer:
A bot sends the code 236. X = 236.
236 > 128. The bot is on team 2 (a predator). 236 - 128 = 108.
108 > 64. The bot is stationary (perhaps a venus fly-trap). 108 - 64 = 44.
44 > 32. The bot has no light sensor. 44 - 32 = 12.
12 <= 16. The bot has one or more touch sensors. 12 - 0 = 12.
12 > 8. The bot only has one RCX. 12 - 8 = 4.
4 <= 4. The bot is sending the message as a 'ping', or question. 4 - 0 = 4.
X now equals 4. This means the bot is bot #4 on the predator team.
So, you've got a stationary predator without light sensors, but with touch
sensors, running on a single RCX, that is requesting information from the
world in general (perhaps another predator, upon hearing this, will respond
with an 'answer' to inform Predator #4 that it is in the vicinity -- or,
upon hearing this request, a prey bot might use a flashing light (being
aware that Predator #4 has no light sensors) to notify its teammates of its
location.) A little far-fetched, but with enough forethought, a lot can be
achieved using communication this basic.
Simplistic, yes, but it shows that you can jam a lot of data into one
number. And sending a sequence of IR transmissions is also a fairly easy
process, to allow you a much more complex string of data.
Of course, we've already contemplated bots that purposefully lie... which,
in my mind, is getting quite a bit too close to human for comfort. Could the
polygraph sensor be far behind?
-- Tom Pierce
>
>
> Stefano
>
>
>
>
>
>
>
> +-----------------------------------------------------------+
> | Stefano Franchi |
> | |
> | Department of Philosophy Phone: Off: (650) 723-2192 |
> | Stanford University Home: (650) 497-2812 |
> | Stanford, CA 94305 Fax: (415) 723-0985 |
> | USA |
> | |
> | e-mail: franchi@csli.stanford.edu |
> +-----------------------------------------------------------+
>
>
|
|
Message has 1 Reply:
 | | Re: legOS (-> Interdisciplinary Learning?)
|
| (...) Are there no bitwise operators? The scheme you described would be much easier to implement [and a damn sight faster] using a bitwise and operator [& in C]. Like the concept though... Hoping to get Mindstorms for Christmas when I will turn the (...) (26 years ago, 2-Dec-98, to lugnet.robotics)
|
2 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
