|
[DPRG] Robert's PARTS ORC Report
Subject: [DPRG] Robert's PARTS ORC Report
From: David
dpa at io.isem.smu.edu
Date: Wed May 28 09:41:57 CDT 2008
Howdy,
For those who don't subscribe to the Portland Area Robotics Society list,
here is Robert Scheer's report on their Outdoor Robot Challenge held
last weekend, 25 May 2008.
dpa
---------------------------------------------------------------------------------------------------------------------------
Thanks everyone for the support and interest in this event. Attendance
was better than I expected, although the rain ended things before we
were finished. Apologies for the delayed report.
Summary:
5 robots showed up and attempted at least the 1st challenge. See Mark
Gross's post for a link to pics of the 5. Thanks Mark for the photos!
It is obvious that kit robotics has not yet descended on the outdoor
robotics world - they're all quite different as was seen in Dallas a
couple weeks ago.
Here are the numerical results.
Challenge 1 Results:
Robot Builder Distance Time
Tarzan Robert Scheer 6' 0:55
Intrepid Mark Curry 8' 2:03
Flash Mark Gross 24' 2:08
AD 1519 Pete Skeggs dnf
Safari Larry Geib dnf
Challenge 2 Results:
Robot Builder Distance Time
Intrepid Mark Curry 6'7" 3:05
Challenge 3 Results:
Robot Builder Distance Time
Intrepid Mark Curry 6'5" 2:09
As is obvious, something was amiss with the precision and order you
might expect at a typical PARTS organized event. Basically, we were
trying to give everyone a chance to test and tune, as was done at the
DPRG event, but the rain was gradually increasing. Since Mark Curry had
the only fully functional bot there, besides Tarzan, we ran Intrepid
through each challenge during opportunities when other people were
testing.
Since I was the organizer/judge, I held off running Tarzan until nobody
else could go. We only made it through challenge 1 before the rain made
it basically impossible to go on. My laptop, which is used to change
course files remotely, was sheeting water. Soooo... we adjourned, most
of us heading to a nearby restaurant for lunch, discussion and awards.
Intrepid won the overall honors and was awarded a 32b AVR evaluation kit
and a well deserved Certificate of Achievement. Everyone who brought a
robot received an award choice. Mark Gross missed out on that since he
had duties elsewhere but we'll make it up to him later.
Comments about robots:
Intrepid was particularly impressive to me, achieving good accuracy with
an Ackermann platform using a motor encoder for distance and a custom
compass setup with Kalman filter for robust heading determination. Even
though Intrepid has been mainly developed for RoboMagellan, it's obvious
that Mark has put a lot of effort into the basics. I know that Mark
only spent some time over a day or two getting Intrepid ready for the
PARTS OC. It's nice to see a well-designed, general-purpose platform.
Unfortunately, you don't get to see much of the guts of Intrepid in the
photos due to its rain coat, but it's the 5th photo in Mark Gross' post.
Flash came out of the closet for this event with an hour or two of prep
also. Flash is the 4th photo in Mark's post. The NiCd battery packs
were the problem here. They were just worn out and couldn't make it
through more than challenge #1. I believe the 24ft distance achieved by
Flash was limited by the battery giving out, so we really don't know how
well it would have done.
AD 1519 was another hastily prepared entry, this one is the 2nd photo
and that's Pete Skeggs shown programming. You can't quite see the
plastic wrap over his laptop keeping off the rain. AD 1519 was the year
that Magellan set sail so here again is another RoboMagellan platform.
This platform moves very smoothly and would obviously be a very precise
navigator if Pete gets the time to work on it, but anyone who knows Pete
knows he's a bit busy to spare much time and it's been on the shelf for
a couple years. There was some bad luck in the last week or so that
Pete has been prepping AD 1519, including the mini-ITX controller not
communicating over it's serial port and the laptop crashing, needing
replacement. The laptop in the photo was being used to control the
robot over 2 USB lines (RS232 and I2C functions).
Larry Geib's Safari started life as a joke but shows good promise.
Larry showed up at the May 3rd PARTS mtg with this thing, just received
from an Ebay auction I think. It's the top photo in Mark's post. So it
was pretty amazing that it could even move by itself after a couple
weeks of work. Unfortunately it wasn't quite able to remember what to
do at the starting line (maybe due to being soaked), but another week
and it would probably be driving Larry to the event.
Tarzan was ready for the event, although it wouldn't have done well on
#4, but was ready to try. Testing earlier had shown 2-3ft errors on the
same field for challenges 1-3 so it's weird that only managed 6ft in
challenge 1. The error was mostly distance measurement so it might have
been something to do with wheel slip. I didn't really care if Tarzan
competed anyway; this was more an effort to promote the event and
encourage people to start building outdoor robots.
More comments and analysis to come forthwith.
Stay tuned.
Thanks again everyone!
- Robert
More information about the DPRG mailing list
|
