[Noisebridge-discuss] Noise-bot 2? (was "RAM me please")
Jake
jake at spaz.org
Thu Feb 17 10:36:42 UTC 2011
ouch, that video was traumatic for several reasons.
this is however very impressive considering it was done in 1997. Of
course hardware has come a very very long way since then (MIT/DIdeas
"Cheap Vision Machine" is a 50 MFLOP digital signal processor with 256K
words of SRAM and a memory-mapped RGB frame grabber).
The question is, can our programmers do what was done then? I think we
can, and it doesn't matter what hardware we use.
I would of course love to build more robots, and I can (and I have the
parts to do so) but I would not want to divert our programmers from the
goal of making robotkind more interesting, to making them more prolific.
Regarding the nimbility of the robot, I realize that the thing has
demonstrated extreme clumsiness in the past, but the python script
governing those movements (controlled through an SSH connection) is almost
deprecated. A new system is being developed by Lilia which controls
direction and velocity (technically rotational velocity and forward
velocity) with a real number, and runs in the background and can be passed
updated values from whatever program is controlling the robots desires.
As I was able to "dock" the robot to its electrical outlet using only its
joystick the other day, I know that it will be quite nimble when
controlled through proper software. As for its size, it will remind us to
keep the space wheelchair-accessible.
Jared, you mention that you want to help with the software side of a
robot; your opportunity is here. Lilia is working on certain things
already but I can steer you toward areas of need depending on what you're
good at, and we can do some interesting things i'm sure.
As for 360 degree sonar and/or bump sensors, we are working on a
bump-skirt right now, and putting sonar on this robot would be a lot
easier than building a new robot and putting sonar onto it.
So... I think we should use the robot we already have and continue to
improve its software.
Once we get enough robot software working to need to branch our code, I
will be very excited to build another one. I have plenty of parts.
-jake
On Thu, 17 Feb 2011, Jared Dunne wrote:
> Jake-
>
> I don't know if you were seriously considering building a new robot or
> not. But it got me thinking
> about if Noisebridge collectively could design and build a more nimble
> robot? It brought to mind robots I used in some robotics classes at
> Northwestern named "Hack" & "Kludge". heh.
>
> Here's an short paper on them:
> http://www.aaai.org/Papers/AAAI/1997/AAAI97-122.pdf
>
> Here's a video of Kludge playing fetch:
> http://www.cs.northwestern.edu/~ian/kludge-with-kids.mpg
>
> It has some of the specs in the writeup but the key features about
> that robot that I found cool:
> - It's roughly the size of a medium sized trash can.
> - It can rotate in place to the right and left, and drive forward and backward.
> - It has 360' sonar sensors to avoid getting too close to things.
> - It has 360' bump sensors to prevent further crushing if impact does happen.
> - It has a forward facing surveillance camera for eyes, but we could
> replace that with the Kinect in our design.
> - It was controlled with an on board laptop similar to MC Hawking.
>
> I alone don't have the mastery of skills needed in the electrical and
> fabrication department, but I would love to help with the software
> side of things and also assist with building/design as much as
> possible. I feel that the skills needed to make something like this
> happen exist within the Noisebridge community. Is there serious
> interest though?
>
> Curious,
> Jared-
>
> On Tue, Feb 15, 2011 at 4:18 PM, Jake <jake at spaz.org> wrote:
>> Unfortunately the PCI7E mainboard which is currently in the robot is only
>> compatible with up to 1GB ram cards. I assume that putting in a 2GB card
>> would either not work, or only come up as 1GB.
>>
>> I guess I should weigh the option of a laptop as opposed to the mini-ITX it
>> currently uses, but I tried a laptop before and there were issues with the
>> power button and stuff like that. Also I would have to migrate away form a
>> parallel port (i assume) which we are using for the chair control.
>>
>> I do have a 2.5" SATA disk though, and Andy has one as well, so perhaps we
>> should just make another robot? A Russian one perhaps? I can certainly get
>> another wheelchair...
>>
>> On Tue, 15 Feb 2011, Dr. Jesus wrote:
>>
>>> On Tue, Feb 15, 2011 at 3:04 AM, Jake <jake at spaz.org> wrote:
>>>>
>>>> dear everyone
>>>>
>>>> the robot is hungry and it wants chips.
>>>>
>>>> RAM chips!! specifically, one ram card of one Gigabyte, DDR2 533 (it
>>>> only
>>>> has one slot and it already has a 512 in that slot) It has to be a
>>>> 240-pin
>>>> type, with a single notch in the bottom. (not laptop memory, it's desktop
>>>> memory we're talking about)
>>>>
>>>> I don't know what DDR2 or 533 means, but if you have a 240-pin 1GB memory
>>>> card, I would love to try it and i'll give it back if it doesn't work.
>>>
>>> I've got some 2GB DDR2 DIMMs. I'll bring them tonight before the meeting.
>>>
>>>> Also looking for a mini-ITX motherboard which is hella fast. This one
>>>> uses a VIA C7 and we're trying to do some heavy 3D imagining using the
>>>> Kinect so the thing can imagine where it is.
>>>
>>> I've got a few laptops you could use. They all have Core 2 Duo chips
>>> in them, but they need 2.5" SATA disks. Do you have any?
>>
>> _______________________________________________
>> Noisebridge-discuss mailing list
>> Noisebridge-discuss at lists.noisebridge.net
>> https://www.noisebridge.net/mailman/listinfo/noisebridge-discuss
>>
>>
>
More information about the Noisebridge-discuss
mailing list