perm filename WISC.L[1,VDS] blob sn#095667
filedate 1974-04-03 generic text, type C, neo UTF8
COMMENT ⊗ VALID 00002 PAGES
C REC PAGE DESCRIPTION
C00002 00002 \\M1BDR25\M2SIGN57\M3NGR25\M4NGR20\F2\CVICARM
\F3\C154 EAST DANA STREET
\CMOUNTAIN VIEW, CA. 94041
\F1\CApril 2, 1974
Professor Richard A. Northouse
Dept. of Electrical Engineering
University of Wisconsin-Milwaukee
Milwaukee, Wisconsin 53201
Dear Prof. Northouse:
I am enclosing some literature on the arms I have designed
and developed. In addition, I am sending in a separte envelope a few
recent reports which cover some of the arm project work at Stanford
There are two basic arm designs which are in current use.
The Stanford type arm is an improved version of the all electric arm
used there since 1969. Several copies have been built and are in use
or will soon be in use at several other research centers around the
country. JPL redesigned the arm slightly to fit their special
application which called for the arm to be mounted on a vehicle and
work on the ground, but the mechanical details and configuration are
essentially the same.
As I mentioned, the selling price of this arm is not yet
fixed, but I think it will be pegged at about $15,000. This includes
a hand, a manual controller, the power supply, servo amplifiers,
brake drivers. All you need is an 8 channel DAC and at least a 16
channel A/D with at least 12 bits resolution. Also, 8 digital
channels are required to drive the brakes. I am redesigning the arm
to allow the use of commercial optical encoders or resolvers. This
will both increase the accuracy and resolution by a factor of two or
so. As for individuals building the arm, I am not opposed to it, but
I have spent a lot of time already telling people how to build the
arm, and then helping with the debugging, so I am not too
enthusiastic with the prospect of taking on another obligation.
There are lots of groups such as yours, who have machine shops,
fully staffed and equipped, but without much work. I have been
thinking of making a kit out of the arm for groups such as yours. I
will supply the drawings for all the necessary machined parts, and I
will also supply all the purchaced components, some pre-assembled
into easy to install assemblies. You would then just make the
machined parts and assemble the arm according to a set of
instructions I plan to make up. I estimate that the kit would cost
about $5,000 to $6,000 plus about 500 man-hours and $250 worth of
material to make the parts and assemble the arm- exclusive of
electronics(estimated at $600 plus 100 hours of labor).
The M.I.T. model arm is much newer. This is a result of the
Mini-Robot project at M.I.T. (part of the A-I Lab.). I have already
delivered the first arm to M.I.T. They have it interfaced to a PDP
11/40 and are developing programs for it. They will get two more
arms in July. I intend to make about 20 of these small arms this
year, at a University price of about $5600 for the arm, a hand, some
electronics and a simple manual controller. Because of the large
number of custom made components in this arm, it is only practical to
make this arm in quantity, so I am not giving out plans, nor thinking
of offering this arm as a kit.
Here are a few names of people who are also thinking of
starting an arm related project of some sort.
Dr. Robert Chen
Coordinated Sciences Lab.
University of Illinois
Urbana, Ill. 61801
Dr. Y.S. Luh
Department of Electrical Engineering
Lafayette, Indiana 47907
Dr. Merill Ebner
Dean of Engineering
Boston, Mass. 02215
I mentioned a few interesting reports produced by other groups.
Here are two, but there are more published by both organizations listed.
"A Mechanical Arm Control System"- A.I. Memo #301 - Artificial Intelligence
Laboratory; M.I.T. ; 545 Technology Square; Cambridge, Ma. 02139. Also,
NASA Technical Memorandum # 33-669 , "Robot Arn Dynamics and Control"; by
Dr. A. K. Bejczy; Jet Propulsion Laboratory; Pasadena, Ca.
Look over the information supplied, and if you have any
furthur questions, let me know.\.