Re: Dropping marker signals?

[Tim Vanderhoek <vanderh@ecf.utoronto.ca>]

This email summarises the relevant discussion that took place Jan. 4
on this thread.

Unanswered questions:  1) What sensor range can we assume in our analysis?

Undiscussed: 1) Viability of having our robot also use directional
                sensing of radio signal emitted from enemy in addition
                to using directional sensing of IR signal emitted from enemy.

After discussion of deployed beacons, the following three possible
uses+implementations were considered:

A) "Double IR method"

   Deployed beacons are equipped with IR transmitters and receivers.

   These beacons can act merely as signal-amplifiers/rebroadcasters (see
below for method of overcoming the "overlapping pulses" problem).
As rebroadcasters, this would only require a single carrier frequency
and wavelength for the IR.

   Alternatively, different carrier signals or IR wavelengths could be
used for each deployed beacon (or, at least, different for deployed
beacons and enemy-mounted beacon).  This simplifies transmission of
simple "yes, enemy is nearby" information and makes realistic the
transmission of additional information (eg. "Yes, enemy just ran over
my weight detector", or "enemy is in direction XY of me").

cons: - There is significant concern that it would be difficult to make
	the deployed beacons tall enough.  Cindy has proposed a
	solution, but it is not entirely trivial.
      - If we are limited to a single IR carrier frequency, there is
	some additional complexity in the necessary modulation
	circuits...my first reaction is that this additional
	complexity is not too considerable.
pros: - we're going to need to build IR circuits anyways, so building
	a few extra will be nothing new, no extra research involved,
	just more parts
      - can also be used to ensure our internal computerized map
        remains accurate

B) "IR-radio method"

    The deployed beacons are equiped with IR receptors and radio
transmitters.

cons: - We don't know how to build a radio transmitter/receiver pair
      - Height-of-sensor problem, as in "Double IR Method"
pros: - Can fairly reliably transmit to our main robot the exact
        direction of the enemy robot relative to the deployed beacon, NO
        MATTER WHERE OUR ROBOT IS ON THE PLAYING FIELD!  This can be used
        deal with reflective surfaces, among other things.

C) "Double Radio Method"

    The deployed beacons are equipped with radio transmitters and
receivers.  The enemy-mounted beacon is also equipped with a radio
transmitter.  The enemy-mounted radio transmitter is powered at such a
level that the deployed-beacon-receivers will only detect the
enemy-mounted transmitter when the transmitter is within a known close
radius of the deployed beacon.  The deployed-beacon-transmitter is
powered at such a level that it can contact our main robot no matter
where our robot is located on the playing field.

    The above produces a simple "yes, enemy is within X radius of me"
signal.  The resolution of this can be increased by transmitting
different radio frequences from the enemy-mounted transmitters, each
powered for a different detection radius.

cons: - We don't know how to build a radio transmitter/receiver pair.
pros: - A very reliable proximity detector --- the "Radio-IR Method"
        is not an effective proximity detector if there happens to be a
        barricade between the deployed beacon and the enemy.  The "Double
        Radio Method" won't even always be able to contact our main robot.

What other pros and cons can you think of?


On Fri, Dec 31, 1999 at 01:23:50PM -0500, Tim Vanderhoek wrote:
>
> I'd be much happier if I were convinced that the speed feedback will
> always be highly accurate.  Certainly it's possible to have highly

The general concensus seems to be "We'll just have to deal with
innaccuracy."


> I'm still thinking about a good way to eject the beacons.  This is
> Cindy's area.  :-)  Propose a method, Cindy.  :)

Method proposed: Jack-in-the-box beacons.  Beacon dropped, after time
                 T, spring unloads and beacons raises itself to height Y.o
                 After brief conversation, this appears reasonably
                 workable.


> > Direct signal amplification may confuse our robot (since it will be
> > possible for the robot to sense 2 IR signals in 2 directions at the same
> > time). 
> 
> Not true.  If the robot is receiving two (or more) IR signals from
> two (or more) directions, it can cross-check these directions with

Additionally, to avoid problems where 2 IR signals turn pulsed IR into
constant IR (which then gets filtered-out), it is necessary to use a
slightly abnormal pulse: the pulse should spend several times lambda
turned off for every time lambda it is turned on.


-- 
Signature withheld by request of author.