To LUGNET HomepageTo LUGNET News HomepageTo LUGNET Guide Homepage
 Help on Searching
 
Post new message to lugnet.roboticsOpen lugnet.robotics in your NNTP NewsreaderTo LUGNET News Traffic PageSign In (Members)
 Robotics / 22624
22623  |  22625
Subject: 
Re: challenge: LEGO copier
Newsgroups: 
lugnet.robotics
Date: 
Wed, 4 Aug 2004 23:07:22 GMT
Original-From: 
T. Alexander Popiel <popiel@wolfskeep.^antispam^com>
Viewed: 
911 times
  
In message:  <joe-5EB4C6.14411504082004@lugnet.com>
             Joe Strout <joe@strout.net> writes:
In article <I1xouE.LpJ@lugnet.com>,
Mark Tarrabain <markt@SPAMBLOCK.lynx.net> wrote:

Actually, he said "flat".  So I would assume that he meant an assortment
of bricks simply stuck to one baseplate, sort of like a mosaic.

No, I meant stuck to each other.

OK, just to clarify: assume an X-Y-Z space with studs sticking up in
the Z direction.  You mean a model that's only one stud thick in X,
arbitrarily wide in Y, and arbitrarily high in Z?

Yes, that would be a much simpler problem than what I was envisioning,
which was X and Y sizes arbitrary, with two interlocking layers in Z
(where the interlocking of pieces extending in either X or Y would
provide the structural integrity of the model).

For what I think you're suggesting, though, it actually would be
fairly simple: have a flat feeder bed that you lay the model down in.
This is gripped by large wheels and rolled (in the model's Z direction)
to a disassembly arm.  The disassembler is on traverse rack as wide
as your maximum Y.  It has a gripper which approaches from both sides
in X.  Starting at one side, with gripper closed traverse until you
hit the first brick.  Then open gripper, move over one stud, then close
gripper on the brick.  Angle gripper up to pop the brick off the model,
then bring in a finger along the Y to determine the size of the brick.
Expose the brick to a light sensor to determine the color, then sort it
into the appropriate parts bin.  Then close the gripper and traverse-scan
for the next brick, until all bricks in the row have been identified and
removed.  Then wheel the model in a bit further, and scan the next row.
This would require 5 degrees of freedom (1 for wheel feed, 1 for traverse,
1 for gripper open/close, 1 for detach motion, 1 for Y-size finger) and
4+ inputs (1 touch for Z position of leading edge of model, 1 touch for
gripper sideswipes brick, 1 touch for y-size finger, and 1 light sensor,
plus any needed for kinesthesia).

Then, on the other side of the parts bins, you have a gripper which can
extract from the bin, traverse to the appropriate point, then attach to
the model being built.  This should require another 5 degrees of freedom
(grab, pull, flip, traverse, and model feed), plush kinesthetic inputs.

Doable, but I count at least 4 RCXes in there.

- Alex



Message has 1 Reply:
  Re: challenge: LEGO copier
 
(...) Yes, but not arbitrarily -- there would be some limit in width (Y) and height (Z), in order for the model to fit into the machine. (...) Yes, that's a much harder problem, which is why I was thinking to limit it to a thickness of one in X. (...) (20 years ago, 5-Aug-04, to lugnet.robotics)

Message is in Reply To:
  Re: challenge: LEGO copier
 
(...) No, I meant stuck to each other. (...) That's what I had in mind too. (...) I was assuming the use of a camera, not the LEGO light sensor. But using the light sensor is a clever idea. (...) Here I expose my complete cluelessness with regard to (...) (20 years ago, 4-Aug-04, to lugnet.robotics)

22 Messages in This Thread:









Entire Thread on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact

This Message and its Replies on One Page:
Nested:  All | Brief | Compact | Dots
Linear:  All | Brief | Compact
    

Custom Search

©2005 LUGNET. All rights reserved. - hosted by steinbruch.info GbR