[IAEP] XO robotics

[Yama Ploskonka]

1) I wouldn't say better... rather, complementary, and certainly 
cheaper. Visiting the Butiá pages, the only picture I see showing an MCU 
http://www.fing.edu.uy/inco/proyectos/butia/images/pistaButia.jpg is 
showing an Arduino. Add a motor driver, and we are well above $30, plus 
shipping. The USBButiá board is maybe cheaper IF done in quantity by 
experts (then add labor).
MSP430 + (L293D OR some darlington array) can be "free" if you get them 
as samples from TI, or less than $5 when purchased, /plus shipping/, the 
old bane. the advantage of using a darlington driver is that then you 
may use plain DC motors, which can be free if lucky with old electronic 
parts (beautiful gear system available in old CDROM drives)

2) yop - the XO "drives" the vehicle with the MSP430 option also. Now, I 
put quote marks as I have no idea - yet - on how to send data direct 
realtime from the XO to the robot, bypassing the MCU. What seems to be 
happening is that Butiá depends on sending code/program to the Arduino, 
and the the 'duino does the brains of the robot. (¿Butiás, me corrigen, 
por favor?). So yes, you may send code to the 430, and then same.
Putting the XO on top of the robot appears to me as a cute photo op 
thing, not a necessity. It would be amazingly cool if the XO camera 
would capture images that real time were processed by the XO, or man a 
sonar sensor, or something. So far I don't think this is happening, and 
the XO on the platform feels like a /pour la gallerie/ innocent gimmick.
Bottom line: control, it's a software issue. I am coding something that 
will allow LOGO-like instructions to be the input, instead of complex-er 
C, but not there yet

3) Oh! but LEGO has a good marketing budget, and I may be mistaken, have 
already been purchased by the brilliant UY education administrators for 
every single middle/high school.
Children have no ideas what is going on anywhere, so I do not worry 
about that (A LEGO driver brick is as much magical and as an MCU).
I do worry about the cost: A LEGO is assured to be carefully kept under 
lock and key, and only safe students and teachers allowed to touch it, 
you can buy a couple hundred MCUs for the cost of a single LEGO brain. A 
430-based solution, and even an Arduino, might have a chance to be lent 
to a kid. A base kit, 430, or ATtiny, say $10 USD including two motors 
and sensors and PCB, could even be purchased by kids or their parents or 
the result of a bake sale. At http://kidbot.org we are working on 
putting such together.
BTW, the Universidad Pontificia of Lima, Perú, and the Universidad 
Católica of Bolivia, I would say the MITs of those respective countries, 
have started degree programs in Mechatronics a couple years ago. Both 
have only ONE semester of actual hands-on robotying, both are Lego only. 
Those professors are so highly trained and respected and bright that 
they must know what they are doing, which means the rest of us must be 
wrong to think other options :-)

4) The choices made by Butiá make sense in some context (UdelaR fing), 
and I genuinely praise and respect their effort, but IMHO are so far 
VERY hard to emulate independently.
IF
1) the PCB were available
2) all parts were locally available, even better as kits, reducing 
shipping/customs issues
I might agree on the "designed to be reproduced by any one".
(their page says they are available locally, but give no details)

What @ http://kidbot.org we are trying to put together is precisely a 
set of materials, parts and such that, more than anything, ARE *easy to 
find parts*. Right now we are VERY busy with the solar vehicle for 
Saturday's race, so this is on hold. On that one: the only part we paid 
for is the solar cells - everything else was scrounged from electronic 
cadavers. Of course we have the benefit that in the USA you can get dead 
printers by simple trashcan-diving, while under certain ideologies you 
have to buy everything. Even then, I ideologically favor plain DC motors 
over complex feedback ones.

I am 100% convinced that the MSP430 could, potentially, provide a 
cheaper, easier to *make* robotics platform, eventually. However, I 
would want to see that as a collaboration with Butiá rather than an 
"instead". I have not contacted them yet on this, as mentioned above I 
have bee busy with the other project (actually mostly with my patent. 
BTW, as of yesterday, I am "patent pending", yeeeha!).

Yama


On 09/27/2012 08:17 AM, Tony Anderson wrote:
> Hi,
>
> I am wondering how this TI MSP430 running on XO 1 - Robotics! relates 
> to the Butia project in Uruguay. Is this a better solution?
>
> What appeals to me the most about Butia is that it enable the child to 
> its own XO driving the vehicle. Can this be done with the MSP430?
>
> In Uruguay there was a demonstration of the Lego robot - pathetic! It 
> was tied to an umbilical cord. It was pre-built so the children had no 
> idea of how it worked or what was going on inside. And, of course, it 
> is frightfully expensive.
>
> The Butia project has developed a control board which is designed to 
> be reproduced by any one. I am hoping that someone in SF will 
> undertake to build the Butia kit. Could the MSP430 provide a cheaper 
> and easier to build control board for Butia?
>
> Yours,
>
> Tony
> _______________________________________________
> IAEP -- It's An Education Project (not a laptop project!)
> IAEP at lists.sugarlabs.org
> http://lists.sugarlabs.org/listinfo/iaep

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/iaep/attachments/20120927/9bc735d4/attachment.html>