COMMENT ⊗   VALID 00002 PAGES
C REC  PAGE   DESCRIPTION
C00001 00001
C00002 00002	\magnification\magstephalf
C00008 ENDMK
C⊗;
\magnification\magstephalf
\parskip=\medskipamount
\chapter{14}{Research Applied to National Needs}

The  National  Science  Foundation  has established a program
called RANN (Research Applied to National Needs) the idea of which is
to  orient  science  to getting results that meet national needs.  In
the  current  climate  of  opinion,  these  needs  are  regarded   as
eliminating pollution, eliminating poverty, improving education, etc.
A recent article in {\it Science} spells this  out  a  bit  and  gives  the
current political situation.

Ever  since  the AI Project made its statement in May 1970, I
have been thinking about how we  might  do  some  applied  work  with
civilian goals.  This has been difficult for two reasons:  first, our
work is, in the main, not ready to be  applied, and secondly, it  seemed
difficult to get financial support.

I  have  finally  reached the conclusion that if we are to do
civilian applied work, we must be prepared to venture  far  from  our
habitual  activities  and  work  with people with whom we are not yet
acquainted.  I would like to start a discussion in the project  aimed
at getting ideas about what we might undertake and to start it off, I
shall mention some possible projects:

\item{1.}Computer controlled or more  generally  remote  controlled
tunneling.

\noindent The proposal is  to  control  tunneling  machinery  from  the
surface using television and other sensors underground. The object is
that  there  should  be  no  people  in  the  tunnel  during   normal
operations.   While  the  main  work of tunneling is done by machines
now, probably there are auxiliary activities now  done  manually,  so
some new machines and ways of using them would have to be developed.

Tunneling  is  dangerous  (17  people  were  just  killed  in
Southern California), a large part of the expense goes  into  keeping
the environment habitable by unprotected humans with minimal success,
tunneling is often done under pressure in which  case  much  time  is
lost  coming  in  and out of pressure, tunnels are big enough jobs so
that some ponderousness in set-up is tolerable, and  money  might  be
saved and tunnels might be completed faster.

If we undertook this, we would have to work with construction
machinery makers the mechanical and civil engineers at Stanford,  and
a  construction firm.  The idea would be to make experimental tunnels
at first rather than to introduce new methods into an  existing  job.
(An  experimental  tunnel would be unprecedented for the construction
industry which only does research in connection with paying projects.

Our contribution  would  be  in  the  areas  of  programming,
interfacing,  human  engineering,  making the machinery controllable,
and getting away from habits of the construction industry.

\section{14.1}{The system envisaged would include the following:}

\item{1.} A control room full of displays including direct  TV  and
also  computer  output.   The  computer  might  be  remote.   In  the
experimental set-up, it could be our PDP-10.

\item{2.}People would go into the tunnel rarely and perhaps only in
protective suits with breathing apparatus.
\vfill\eject