COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002		The preparation of drawings to control the construction of buildings
C00009 ENDMK
C⊗;
	The preparation of drawings to control the construction of buildings
and the production of smaller objects is an important source of delay and
expense.  Besides drawings themselves, bills of material, instructions
for numerically controlled machine tools and automatic wiring machines are
becoming of increasing importance.  All these normally involve a different
group of people from the engineers and architects who design the products.
Modification of drawings once produced is also very expensive especially
since the modifications have to be made by different people than originally
made them.  Anything that can reduce the cost and delay of drawing will
increase the promptness with which society can respond to its technological
problems.  The current energy crisis illustrates the fact that we often
don't recognize problems until they are so acute that fast action is
required.

	Many computer aids to design have been developed since the late
1950s and have met with a mixed reception.  Some have met with complete
acceptance and are the standard way of doing things in particular
industries.  For example, computer aids to the preparation of integrated
circuit masks are in wide use.  Other proposals have not proved cost-effective,
because the equipment was too expensive or because the system was too
inflexible.  Our own experience with computer aids to the production
of drawings has ben good.  The system developed here by Richard Helliwell
for making logic diagrams, printed circuit board layouts, instructions for
automatic wiring machines and board drilling machines is in continuous use
here and parts of it have been adopted by MIT and by Digital Equipment
Corporation.  Part of the success of the system is based on the fact that
it operates in the context of a general time-sharing system  with
all of its facilities and doesn't represent an independent investment in
hardware.

	Our proposed research will advance the art in the following respects:

	1. The direct description of three-dimensional objects will be
more easier and more flexible than the present technique of making two
dimensional drawings.  Such descriptions will require less change
when design is changed and can produce whatever views are required
when requirements change.

	2. The three-dimensional descriptions will be more convenient for
generating information for controlling machine tools and automatic
assembly systems.

	3. The resulting system will run on a standard time-sharing system
for the PDP-10 (certainly on TENEX and probably on TOPS-10 also).  An inducstrial
user of the system will require only suitable display terminals which
can be connected to the computer by telephone.  This will reduce by an
order of magnitude the capital cost of computer aided design, since most
of the present systems run on dedicated computers.

	We are not yet in communication with potential users of this system,
but we anticipate interest comparable to that already shown in our
logic and circuit design system.


	This work is related to other activities of the Stanford Artificial
Intelligence Laboratory, especially to the work in automated assembly
which needs to start from descriptions of the objects being assembled.
The ability of the proposed system to handle motions of the objects as
well as static descriptions will aid in the development of the 
"automation assembly language".  It is also related to a proposed
project aimed at an automated machine shop.


	The Stanford Artificial Intelligence Laboratory has a management
structure suited to this kind of project.  Both the Director, John McCarthy,
and the Executive Officer, Lester Earnest, have experience in managing
such projects and a special interest in this one.


	Successful completion of the research will make available programs
that will make design faster and cheaper.