Subject:
|
RE: Any Comments on "Light Bricks?"
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 15 Sep 1999 19:50:10 GMT
|
Viewed:
|
729 times
|
| |
|
|
Frank:
I think you will find that the optics on the front of the light sensor brick are
too wide angle to do much in the way of beacon detection. You would probably
need to put at least a field lens in front of the brick, and perhaps some
tubular shielding to improve the discrimination of the sensor. I also remember
from the postings that the sensor is more sensitive to the IR end of the
spectrum, which might lead you to different light sources. Laser pointers are
cheap, but would be too finely collimated to find, IR diodes are also cheap, and
could be battery powered to spread around the room or on top of 'bots.
I am working on a navigational concept right now that uses the "beacon" idea,
but I haven't gotten it far enough along to put up pictures (and spend even more
money on technic parts). I am using a CdS cell, which varies resistance with
light intensity. I have put the sensor at the focal point of a parabolic mirror
(solar cigarette lighter from an old, old, robot kit). The mirror seems to
improve the discrimination of the sensor, bright objects are still bright, and
darker objects, walls, ceiling, etc are "darker". I have put the mirror on a
two axis mount, one rotating from horizontal to straight up, and the other for
panning side to side. Note: This second axis also controls the front wheel for
steering, so the 'bot is steering toward where ever the mirror is pointing.
Both axes are significantly geared down, and have rotational sensors. I haven't
calibrated it yet, but my guess is I have resolution of about a degree, which
should be fine for getting around the office.
My idea is to put the IR transceiver from the computer at the base of the
"brightest star in the sky", so that when the "bot finds it, (with the rcx brick
mounted facing forward) it can also tell me where its been, and what its been
doing. I think I can also use the CdS for taking pictures, using the same scan
routine, but logging to the data log instead of looking for a sensor value from
a light.
With two lights of different intensities (and therefore known resistive values)
I think I will be able to do "Celestial Navigation". The lack of trig functions
is a pain, I am going to try using a bunch of constants ( for tan(angle) 0 to 90
) to be able to figure out where I am.
A rather long reply, my first post. I have had this rcx kit for ten days now,
and I am absolutely hopelessly hooked.
Jeff
Jeffrey.hazen@northmill.net
|
|
1 Message in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
