Subject:
|
Automation Project Proposal
|
Newsgroups:
|
lugnet.org.ca.rtltoronto
|
Date:
|
Sat, 8 Jan 2005 15:40:06 GMT
|
Viewed:
|
591 times
|
| |
|
|
I know everyone's gung ho on balls, but I'll bring this back to the table
for a thought:
A long time ago (over 3 years ago), I started a little project to see if I
could build an automated assembly carousel / high speed assembly dial to
build a simple LEGO construction. I made a simple duck, with a 2x1 inverted
slope brick for the body and tail, 2x1 plate for a beak, and 1x1 brick for a
head. I'm *still* trying to build a dial-based automation to do this. I've
somewhat lost interest in it for a while since I'm currently working on an
8-station high speed assembly dial, with ultrasonic welding, leak testing,
and 3-parts assembly (at work) and it's too much fun :)
Anyway, when I talked about it with Calum and Rob, the 3 of us thought it
would make a great project to do as a group. Build modules of automation to
a specific size. Use pallets for proper queueing and buffering between
modules (pallet based automation is simple and extremly flexible, it is the
easiest way to integrate several suppliers automatiuon together, as far as
I'm concerned). We could design a more complex part to be assembled, and
each person would be assigned an input state and output state for the part.
So, using the duck for example, I might have the first task of loading the
body piece into the pallet, then passing it off to Dave's module. Dave's
module places and presses-on the beak. Then Dave releases the pallet to
Ka-On, who has to add the head. Ka-On's module passes it to Wayne's, where
he has to remove the ducks and stack them using a cartesian gantry and
palletizing algorythm into an outfeed system, or something.
I'd imagine each module would have at least 2 RCX's - probably one alone for
pallet handling and at least one for performing some kind of task. So it's
obviously a multi-person project, I've only got 3 RCX's and I've only done
one 3-RCX automation project. But we shouldnt' even need any communication
between cells if we do it right. As long as the unit runs starved and
pallets are introduced at some kind of regular interval and there's no one
module that takes a huge amount of time we'd be fine. Otherwise we could
clock-pulse it with a master RCX sending a trigger pulse over IR, something
like that.
Iain
|
|
Message has 1 Reply:
 | | Re: Automation Project Proposal
|
| (...) You know me, I'm all for this kind of project. The marbles are interesting, but not nearly as cool as actually making something out of lego. This is pretty much an independant project though. I doubt there will ever be many people who are (...) (20 years ago, 9-Jan-05, to lugnet.org.ca.rtltoronto)
|
14 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
