<HTML><HEAD></HEAD>
<BODY dir=ltr>
<DIV dir=ltr>
<DIV style="FONT-SIZE: 12pt; FONT-FAMILY: 'Calibri'; COLOR: #000000">
<DIV>John,</DIV>
<DIV> It is funny that you ask. The DPRG monthly meeting topic
is being presented by a Hackathon team made up of DPRG members (see dprg.org for
news article). They used an online object identification (training set) as part
of their project. </DIV>
<DIV> </DIV>
<DIV> If I were interpreting current RoboColumbus Plus
rules, I would say that a web based training set app would fail the on-board
test. The reason is because the rules specify specific types of allowed data
transfer (It might also run afoul of the beacon definition). However, I for one
would be ok with adding an off board image identification training set app to
the list of allowable data transfer. The trend seems to be using large web based
object training sets. I do wonder in the case of an orange traffic cone, if a
sufficient training set could not be built and run on one of the available linux
boards (RPi, BBB, STM32F7, Edison, etc...). I think you could identify it by
color for example <grin>.</DIV>
<DIV> </DIV>
<DIV> Make sure you come to the Saturday meeting at DMS. We
could discuss with others. I like your ideas.</DIV>
<DIV> </DIV>
<DIV> As for being stuck in high grass, the robots managed to
navigate a fairly boggy section with high grass during the last contest. I could
see a leaping/bouncing robot that didn’t go too high or offer too much of a
scary situation, and was safe being ok. The rules do not really cover a
ground based robot that occasionally leaves the ground. For example, I don’t
think a robot would be eliminated that left the ground due to an naturally
occurring ramp. </DIV>
<DIV> </DIV>
<DIV>Regards,</DIV>
<DIV>Doug P. </DIV>
<DIV
style='FONT-SIZE: small; TEXT-DECORATION: none; FONT-FAMILY: "Calibri"; FONT-WEIGHT: normal; COLOR: #000000; FONT-STYLE: normal; DISPLAY: inline'>
<DIV style="FONT: 10pt tahoma">
<DIV><FONT size=3 face=Calibri></FONT> </DIV>
<DIV style="BACKGROUND: #f5f5f5">
<DIV style="font-color: black"><B>From:</B> <A title=swindle@compuserve.com>John
Swindle</A> </DIV>
<DIV><B>Sent:</B> Tuesday, December 6, 2016 11:01 AM</DIV>
<DIV><B>To:</B> <A title=paradug@gmail.com>paradug@gmail.com</A> ; <A
title=dprglist@lists.dprg.org>dprglist@lists.dprg.org</A> </DIV>
<DIV><B>Cc:</B> <A title=swindle@compuserve.com>swindle@compuserve.com</A>
</DIV>
<DIV><B>Subject:</B> Re: [Dprglist] RoboColumbus rules
question</DIV></DIV></DIV>
<DIV> </DIV></DIV>
<DIV
style='FONT-SIZE: small; TEXT-DECORATION: none; FONT-FAMILY: "Calibri"; FONT-WEIGHT: normal; COLOR: #000000; FONT-STYLE: normal; DISPLAY: inline'><FONT
color=black size=2 face="Arial, Helvetica, sans-serif">
<DIV>Doug,</DIV>
<DIV> </DIV>
<DIV>A couple more questions:</DIV>
<DIV> </DIV>
<DIV>Does the self-contained definition prohibit the robot from asking a website
a question? If all the decisions are made on-board, and the website is not
merely giving coordinates to the robot, would that be OK? Ezra sent me several
links for agricultural robots and plant identification apps. Good stuff. So, I
wondered about having the robot look at things and ask a website what it was
looking at. The site might reply, "That's an oak tree."</DIV>
<DIV> </DIV>
<DIV>Staying on the ground: So if a robot is not flying, but also is not staying
on the ground, would it be OK? No, not rocket power. But bouncing, maybe with
some propeller lift, but not enough to achieve flight. Just thinking of ways to
avoid getting stuck in mud and high grass.</DIV>
<DIV> </DIV>
<DIV>Thanks!</DIV>
<DIV> </DIV>
<DIV>John Swindle</DIV>
<DIV> </DIV>
<DIV>
<DIV
style='FONT-SIZE: small; TEXT-DECORATION: none; FONT-FAMILY: "Calibri"; FONT-WEIGHT: normal; COLOR: rgb(0,0,0); FONT-STYLE: normal; DISPLAY: inline'><FONT
face="Arial, Helvetica, sans-serif"></FONT><BR></DIV></DIV></FONT></DIV></DIV></DIV></BODY></HTML>