[Robotgroup] Drive a Droid proposal... How to detect the "fence"?

[Doug Emes]

> From: "Eric Lundquist" <roboenator at gmail.com>
> Instead of RF, how about a magnetic trip switch?  I have a roll of the
> magnetic strip material used in making refrigerator magnets.  It is not very
> strong but could probably be used in conjunction with a hall effect switch
> mounted onto the robot similar to line detecting sensors.

The issue here is the pittman motors / NPC 2212's should give off a fair amount
of interference, could you shield the hall effect sensors pointing down in the
three feet to NOT be affected by the drive motors? If you can, then this could
be a nice solution.

Another thought I had was to have a couple IR lasers "beacon" out across the
stanchions, and if one is tripped, you know which side R2 is trying to escape
from, and you could then work on having R2 return to center based on some
algorithm (tbd), including possibly maximizing the distance from 4 sonar
beacons (on in each of the 4 corners).  the advantage of this design is you only
need 4 corner units, the disadvantage is that if you bump or twist one, the
system wont work.  Ah, the fun problems we make and try to solve...

In reality, this is a fantastic problem set not just for the r2
builders, but for anyone
who wants to demonstrate a robot (either RC or autonomous) and be able to
"keep it in a virtual box" that can be moved from convention center floor, to
hotel lobby, to family garage, to you name it.

Doug Emes
DPRG Librarian