Subject:
|
January 2005 Meeting Notes
|
Newsgroups:
|
lugnet.org.us.smart
|
Date:
|
Wed, 2 Feb 2005 19:37:05 GMT
|
Viewed:
|
3000 times
|
| |
|
|
Saturday's SMART meeting went really well! For our fourth anniversary we had
slightly less people than normal, but almost everyone brought a mini-challenge
or other robot. The mini-challenge was to build a robot that could detect a
60-watt lightbulb, and shoot a projectile at that target. It always amazes me
how many different ways there are that people come up with to solve the
problems. This time not only were the mechanical solutions very different, but
obviously so were the algorithms people used. (Often it's hard to tell, but this
time it was easy to see the different search strategies.)
Mechanically the main thing that everyone had in common was some sort of
turntable used for turning the launch platform, and a light sensor to see the
lightbulb. But that's where the similarities ended. Projectiles included a
catapult that launched the small LEGO soccer balls, to axles, to minifigs!
Finally a practical use for minifigs in a LEGO robot! :-)
Some people had their light sensor completely 'bare', and looked only for the
maximum brightness. Others had different types of shrouds. Two examples: One
robot had a shroud only on the left side and bottom; their robot would scan to
the left, and when a there was a sudden jump in brightness, it knew the bulb had
just come into view. After moving up, it scanned down, again to find the sudden
peak. Another robot had a narrow vertical slit used to find the horizontal
position of the bulb, and a tiny round hole to find the vertical position.
Algorithmically, one robot did a complete scan of the environment, remembering
the brightest X & Y coordinate (or I guess yaw and pitch would be more
accurate), returning there at the end. Another divided the area into quarters,
found the brightest quarter, divided that quarter into smaller sections, etc.,
until it had found the light. Several scanned horizontally looking for the
brightest, then vertically, but even there, different methods were used -- a
calibration pass first for highest and lowest values, or just a quick scan for a
local maxima.
Besides the mini-challenge robots, there were a couple of other robots people
brought. Mark brought his crane that he built for the last Crate Contraption.
Someone brought a paper-towel dispenser (the kind you wave at to get a towel),
and someone else brought his version of a Legway that used normal Lego light
sensors. I'm sure there was another one or two I forgot. And a number of people
have been working on jam-proof hoppers for our Crate Contraptions.
After the mini-challenge we talked about the next meeting. It will be at a local
mall (Crossroads Shopping Center), instead of at Digipen. We'll still meet from
2-5, but the entire meeting will be running our mini-challenge robots: an Easter
Egg Hunt. Of course there will still be plenty of time to show and tell, talk to
people, etc. I'll describe the challenge in the next message, but I hope that
everyone will build a robot and bring it to the show. It will be a great time to
show your robots off -- you can bring other robots as well -- and you'll have a
chance to make kids real happy as they watch robots deliver chocolate Easter
Eggs to them.
I've been waiting for the folder with pictures of the meeting to be moderated,
but it's not happening yet. So whenever it finally happens, you'll find the
pictures (and a couple of video clips of some of the robots in action) here:
http://www.brickshelf.com/cgi-bin/gallery.cgi?f=115017
--
David Schilling
|
|
1 Message in This Thread:
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
