Random assortment of thoughts

[Joyce Poon <poonj@ecf.utoronto.ca>]

Hullo,

I just checked out the web page and Malone doesn't have office hrs at 8am
on Mon.  We can either see him tomorrow after 10 or next Wed.  I think
tomorrow may be a better time so we can have more time to think over what
he may tell us.  But I don't have a circuits lab... so it's up to you
guys.

On another note, I also checked out the past designs on the web.  The
following is what I gathered from the abstracts:

a) Sensors
It seems that infrared sensors were the most common, although 1 grp did
mention ultrasonic sensors. IR emitter and photosensors were used
extensively.

b) Electromech
2 wheeled systems employing 2 motors also seemed to be the norm. We should
also discuss the speed & maneuverability (sp?) of the robot.  With the
exception of 1 grp, all robots had some sense of direction in terms of
mvm't (ie going backwards, forwards, turning etc.)  We should also think
about how everything is going to fit onto the robot (how many "levels"
should we have/do we need?) Oh yeah, aluminum & lexan (what's that?) seem
to be a popular choice in terms of material.

c) Interface
The biggest challenge is to design a circuit that can cleanly pick up the
beacon signal (& not some random noise gen by motors and such).  Also, I
think that the power/current surge/draw during the power-up phase may also
have caused problems for a number of grps.

d) Computer
We have to come up with a clever algorithm.  Should we have diff
algorithms for evade & catch or should they just be the opposite of each
other?  Moreover, 1 grp mentioned that it is a lot easier to fix
problems/improve the performance of the robot by changing the software
rather than adding circuits. I think 1 grp (not tag robot though --
Rubic's cube solver) used C++ to code their algo.  A number of grps used
the motorola 6502 board & programmed in assembly.

e) Other things
A lot of people mentioned that more time should be spent on the
integration (not be confused with differentiation... haha... bad joke...
even for an eng sci) stage of design.  We should keep that in mind when we
work on our robot.

Well, I think that's enough design for a day.  Time to do some circuits
problems...

Good luck on cir lab.

Joyce :)