[Gghc-discuss] topic of project

[Erich Heine]

HI Jonathan,

Great questions.  Not all of them have an answer so far, but here is a bit
of a synopsis.  I may go into a bit more detail than is warranted here, but
thats ok, we can then turn it into a blog post for the element 14 site :)

So the general idea at this point is still the same as or application
writeup.  Some things have changed in the short-term though. This GGHC
project has a fairly hard deadline of 5 weeks or so from now. This really
limits what we can realistically do -- just the amount of learning and
designing on our end really affects the scope.  So we have decided on a
somewhat more modular approach. Modular in this case actually has a couple
different meanings:

* Do this project in fully formed steps, so we can at least get /something/
done by the gghc deadline -- the organizers are pretty vague about
definitions of "done" since the time is so short and this is new to them too
:)

* The project results are fairly modular -- e.g. courses around the robot
and courses around the communication and courses around the robot
cooperation can all be separate weeks for some sort of class designed around
this project.

The details on how this actually translates to our tasks is still undecided
at the moment, but we feel this gives us the best chance for being realistic
about this. A further consideration on our part, is that with this staged
approach, we can keep working on this even after the GGHC has ended, because
it is a fun idea with or without a contest.

Here are some of the steps we identified towards building our project. These
aren't final, or may not even make full sense, just a starting point:

* Get the robots working and doing (semi-) autonomous movement -- e.g.
figure 8s
* Get a minimal mesh building software going
* Super simple search and mesh building algorithm -- no autonomy on the
robots part, no swarm or anthythiong
* Work on a distributed, swarm style algorithm for the bots
* document this stuff for reproduction
* test our stuff with real kids in a real learning environment.

And so on. Those last two are really important for both the contest and the
general scope of the project -- if we want this to end up helping teach
people we need to try it out!

As for the more technical questions:

Bot Brains: We are thinking Arduino at the moment, due to it having
momentum, lots of pluggable kits, and lots of free code in the world. This
is not a final decision, there are other good products out there such as PIC
and the TI Launchpad -- mostly we need rapid development capabilities
combined with affordability and proper processing power. If you know this
space, your input would be really appreciated!

Sensors: We are actually looking at the minimum viable robots at the moment.
This mostly means comms, wheels (and wheel control), processor and thats
really it. We are currently looking to get mvp stuff happening in a large
open space, eg parking lot or gym.  Sensors are a fantastic idea, but having
our robots find thier way around a room is a hard problem still subject to
serious research. If you have thoughts on this, or know more than my
paragraph suggests: input please! Sensors are cool and incorporating minimal
sensors would be neat.

Wireless communication: we are looking at zigbee, via xbee modules. These
seem to be the slickest of the available solutions, and we don't have to
write our own meshing on top of the rest of it. Also, they are low power :)
 There really don't seem to be many alternatives at this point either.

Base station: probably just a laptop with linux or windows and an xbee usb
dongle. This is the simplest, stablest develpment platform for now. Future
revisions can change this but this will probably give us the least "get
going" issues.

Software: we plan on leveraging anything that already exists, but writing
new stuff will be part of it too. This at the moment is the biggest unknown
from my POV.


HTH
Erich

On Sun, Mar 27, 2011 at 10:00 PM, Jonathan Manton <jmanton at illinois.edu>wrote:

> So, I found Erich's writeup of the project proposal at
> http://wiki.ucimc.org/bin/view/MakerspaceUrbana/GreatGlobalHackerspaceChallenge.  Is the project described there still the general idea?
>
> If so, do you guys have thoughts about:
> - "brains" for the small robots (e.g., is it an arduino, a PIC, whatever);
> - what sensors would be on the small robots - what "signals" would be
> objective points give out;
> - what type of wireless communication will be used;
> - architecture for the base station (e.g., Linux PC or an embedded platform
> like a BeagleBoard); and
> - will the software be based on something that is already out there, or is
> this from scratch?
>
> I'm particularly interested in the last point, due to the tight timeframes
> involved.
> _______________________________________________
> Gghc-discuss mailing list
> Gghc-discuss at lists.chambana.net
> http://lists.chambana.net/mailman/listinfo/gghc-discuss
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.chambana.net/pipermail/gghc-discuss/attachments/20110328/1a6ca2bd/attachment.html>