[Dprglist] Sample Retrieval vacuum cleaner

[deltagraph at aol.com]

Doug / All,
 "I challenge you to outline an alternate method that is actually superior to vision in performance for this task."
I can think of a inexpensive "1D" sensor that can do a better job than your camera classifying your objects excluding different "paint jobs" and it does not require a super computer. You might have to strain a bit to think what it is or go to a lot of trouble to put objects in with the same mass, but different colors... 

Ron Grant
 TwoDollarLoadCellsForYou.com
PS might want to add that to the exclusion list to _force_ me to abandon my creative thinking and conform to the  vision camp! Ha!  I overheard someone say that they were hoping John Swindle would build his bat bot to solve this problem. John might argue that you are evolving anti-sonar strategy, like the perhaps the Hawkmoth with its anti-sonar "feathers" that might just force bats to re-activate old genes or re-evolve genes for eyes.

Interesting high frequency sound is used to measure / image eyes and all their internal bits, for cataract surgery -- like 10 MHz, would be interesting to learn about benefits of high frequency sound to better image objects / textures... 




 
-----Original Message-----
From: Doug Paradis via DPRGlist <dprglist at lists.dprg.org>
To: John Swindle <swindle at compuserve.com>
Cc: DPRG <dprglist at lists.dprg.org>
Sent: Sat, Jan 19, 2019 11:11 am
Subject: Re: [Dprglist] Sample Retrieval vacuum cleaner

John,    Yes, a shop vac might have a role in a robot retrieving samples, but there many much simpler ways to collect the objects. The one group that I think has substantial experience in collecting items, is Iron Reign. They have built many robots for various FIRST competitions which use a spinning mechanism with a zip tie broom that sweeps items into a hopper. As i stated in our previous message exchange, the rules do not require that the specific objects be retrieved one at a time or be placed on the floor of the home base (only in the home base).
   Also, the rules do not require the hunted objects to be in sight of the home base, only that they be in the search area. There may be obstacles between the home base and the object's location. So in my opinion truly searching for the objects would be a better strategy, opposed to trying to a  "collect them all and let god sort them out" method. We will see what the contestants dream up in November.
   BTW, with the current state of technology available to hobbyists, I think that vision is not only cheaper, but better. I challenge you to outline an alternate method that is actually superior to vision in performance for this task.
  I hope to see you compete. Start thinking about the contest and go for it.
Regards,Doug P.
   
On Fri, Jan 18, 2019 at 11:00 PM John Swindle via DPRGlist <dprglist at lists.dprg.org> wrote:

Doug, et al, Good to see my emails to the list are no longer bounced. My comments about self-driving cars got me on the evil list for a while. (I would still rather be face-planted and rear-ended constantly than have any measurable chance that my car's software would kill me.) A solution to Sample Retrieval was voiced at the January meeting: a leafblower, but since there are no arena walls, that would not work. But a shop vac would. Seems that a statically-guided shop vac would win the contest. As usual, I am missing the point. Thank you for reminding me that vision processing is chosen not because it is good but because it is cheap. I'd forgotten that. Indeed, in my classes, I emphasize that the choice of processor, memory, and expansion bus has less to do with performance: It has most to do with what is cheap. I have a Kirby, but Hoover is better known for cleaning up. Seems like a solution to the contest. No height limit given. Best to you, a very patient person, John Swindle 
_______________________________________________
DPRGlist mailing list
DPRGlist at lists.dprg.org
http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org

_______________________________________________
DPRGlist mailing list
DPRGlist at lists.dprg.org
http://lists.dprg.org/listinfo.cgi/dprglist-dprg.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.dprg.org/pipermail/dprglist-dprg.org/attachments/20190119/853b1db0/attachment-0001.html>