Grex Science Conference

Item 5: Teeny tiny little machines

Some of you remember Japan's "Fifth Generation Computer Project".
It turned out to be a big fizzle, producing very little to market.
It is an example of how large government spending on research projects without
a clear marketing goal may not produce useful results. 

I bring this up now, because Japan has embarked on a new government-sponsored
research project, and I wonder if it will be different this time, or it will
end up the same way. The project is the "Micromachine Technology Project".
This project involves 26 companies with financing from  the Ministry og
International Trade and Industry. It is 5 years into a 10 year program and has
a budget of about $250 million.

They are constructing microscopic machines and tools. There is an article 
in the NY Times "Science Times" about this project. The article features a
description of a 7mm long miniature Toyota automobile with a functioning drive
(front-wheel drive).

By contrast, the US is also spending money on micromachine technology. The
Defense Dept. DARPA is the main sponsor, laying out $35 million in 1995.
US research is different though. It is mostly two-dimensional
photolithography-based silicon, also called microelectromechanical systems
(MEMS).

Japan is not totally ignoring this area, but is focussing on the
miniaturization of more conventional machining processes. Their scientists
argue that the broader approach will allow them to make micro-machines from
a wider variety of metals and ceramics, providing more strength and
versatility.

Japan's prject has three target appicltions -- the pipe inspection machine,
which crawls down the inside of pipes, looking for cracks, medical devies that
can be threaded through blood vessels, and a micro-factory small enough to
fit on a desk, to be used to make miniature parts for watches and cameras or
other micro-machines.

That's interesting, but folks should ALSO keep in mind that that japanese
funnel those funds thru "Kobuns" which are formed of hundreds of
interdependent companies and form a virtual monopoly on a subject, idea
concept, etc..

In the US, you have Feds and judges that frown on the very idea, and the
single attempt of a few years ago was completely ruined before the
Chip-manufacturers got anywhere with the deal..

Meanwhile, DARPA is also getting a small part of the entire pie, and it's
almost exclusively Research, where a Kobun is the entire enchilada:
research, manufacturing, supply, marketing, etc, etc..

Makes competition with them a bit rough on American firms, neh?

The medical instruments sound like a good idea, but I *was* wondering who was
going to drive the mini-truck.  =)

GIven that miniaturization has been one of the greater  forces
behind cost-cutting and enhancing the market value of goods,
I'd say this has someting to it.  After all, we don't use
fourteen inch disk drives any more, do we?

We didn't like the 3" floppies either and 1.8" HD are having a difficult 
time at teh moment.

Remote control. Sarah. You could even drive the little truck.
(But it'll probably be an MD who drives it for medical applications.)

There are many things different about the way the Japanese conduct research
and do business, versus the western way. In this case, I am interested in the
difference in approach -- minimize conventional machines (the Japanes
approach) versus a whole new kind of (mostly 2-d) machine, made by lithography
rather than machining (our way).

Is this a replay of the Japanese Analog HDTV mistake? Or are they  on to
something we shouldn't be ignoring?

It seems like this latest Japanese project has a greater chance of
success because the goals are much more tangible; making tiny machines
that perform useful functions.  Their computer project was relatively
vague and less focused because artificial intelligence is a vague
concept.  The notion of what AI is has changed over the past few
decades, and it continues to change.  Whenever the basic principles
of an AI field are worked out, it ceases to be AI and becomes its
own field, like compilers, expert systems, and neural nets.
 
Tiny machines (also known as "nanotechnology") look like a promising
research area, which will give us more control over the physical
universe.  For example, thousands of these machines could be injected
into a body to perform repairs in situations where surgery is infeasible.

I suspect that there is a more sinister side to that one too.

In a sense we have artificial intelligence now, in the form of expert systems.

It will be a few years before we can build a microscopic machine that has
commercial value, but we can build elementary microscopic mechanical devices.

Nanomachines injected into the bloodstream are further away yet. I doubt we'll
see them in the next 20 years. Yes, I know how fast technilogy is
accelerating. I think people are underestimating the extent of the missing
technology required to get that kind of stuff to work.

• Read the next item
• Go to the menu of items for the Science conference
• Go to the list of conferences
• Go to the opening screen