Article View: comp.org.ieee
Article #478Re: Robotics
From: John B. Nagle
Date: Fri, 16 Dec 1988 03:38
Date: Fri, 16 Dec 1988 03:38
23 lines
1676 bytes
1676 bytes
How did this get in here?
Anyway, assuming that you know where the surfaces are, and just need
to do the position control, the problem is straightforward. See
"Robot Motion - Planning and Control" for the basics. ISBN 0-262-02182-X.
If you only know approximately where the surfaces are, position to a point
known to be off the surface and use force feedback to servo the pen onto
the surface, with the force feedback operating only perpendicular to the
surface. Then write, with the axes parallel to the surface controlled
by the writing algorithm and the axis perpendicular to the surface
controlled by the force servo loop. If you put some vertical compliance
in the pen holder and provide the pen force to the controlling computer,
that should be sufficient for servoing, assuming the pen is suitable
for constant-force dragging across the paper, and not something exotic
like a nibbed pen or a calligraphy brush.
No matter what orientation the pen is in, this should work. You
do all the writing and servoing computations in coordinates relative to
the writing plane, then convert them to absolute coordinates, then
compute the inverse kinematics and issue the commands to the robot.
John Nagle
Technical Details
Message-ID:
<17906@glacier.STANFORD.EDU>Path:
arc1-ca.pugleaf.net!utzoo!utgpu!attcan!uunet!lll-winken!lll-lcc!ames!pasteur!agate!labrea!glacier!jbnReferences:
<5166@bsu-cs.UUCP>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!utzoo!utgpu!attcan!uunet!lll-winken!lll-lcc!ames!pasteur!agate!labrea!glacier!jbn From: jbn@glacier.STANFORD.EDU (John B. Nagle) Newsgroups: comp.org.ieee Subject: Re: Robotics Keywords: Robotics, Path control Message-ID: <17906@glacier.STANFORD.EDU> Date: 16 Dec 88 03:38:14 GMT References: <5166@bsu-cs.UUCP> Reply-To: jbn@glacier.UUCP (John B. Nagle) Organization: Stanford University Lines: 22