Micro-Fabrication Article

From: Lloyd G. Rasmussen (lras@loc.gov)
Date: Thu Apr 03 1997 - 06:57:19 PST


The following is from EE Times, CMP Publications. It reminds me very
much of Tim Cranmer's talk at the technology conference.

                                  [ISMAP]
                                      
    Fujitsu -- World's Fastest Growing Disk Drive Manufacturer Visit the
                              Altera Home Page
                                      
                                  [INLINE]
                                      
                  "Microbots" aid shift to 3-D fabrication
                                      
                              By Sunny Baines
                                      
   LAUSANNE, Switzerland -- Microlithography has been successfully
   adapted to building a variety of micromechanical systems, from airbag
   sensors to micromirror imaging chips. Nevertheless, some researchers
   are growing impatient with the limitations of lithography's inherently
   planar nature.
   
   Moving in new directions in microfabrication, beyond the layered
   structures and limited materials of lithography-based batch
   processing, many scientists are looking to "microbots"--small robots
   that can operate with micron or better precision--to do the
   construction work.
   
   A process that can make mechanical, optical and electronic parts
   separately and then put them together after fabrication will be
   essential for truly three-dimensional microrobotic systems. Several
   groups of researchers--including teams at the University of Tokyo,
   Carnegie-Mellon University in Pittsburgh and the Swiss Federal
   Institute of Technology, here--are looking into the problems
   associated with this emerging technology.
   
   Assembling truly microscopic systems on microfabrication lines will
   push basic robotic functions--sensing, actuation and control--to their
   limits. In sensing, for example, conventional optical machine-vision
   techniques are not good enough for microassembly tasks that require
   nanometer precision, because the wavelength of light only allows
   resolution to about 1 micron. Likewise, traditional robots are simply
   too clumsy to make precise movements at nanometer scales. The actuator
   backlash--the slight oscillation caused by inertia when a robot arm
   stops moving--alone is enough to destroy a delicate micropart.
   
   On top of that, the mechanics of picking and placing components is
   completely different at microscales. Overcoming the force of gravity
   becomes irrelevant compared with dealing with the Van der Waals forces
   between microscopic particles and such other physical effects as
   humidity.
   
   Another barrier to building fully integrated micromachines is
   economic, said Jean-Marc Breguet, a senior researcher at the Institute
   of Micro Engineering, part of the Swiss Federal Institute of
   Technology (EPFL) in Lausanne. "We must consider two different market
   segments," he said: "mass production and small to medium production."
   In mass production, "large investments are acceptable for production
   lines, such as in the IC industry," Breguet said. In small to medium
   production environments, "a pragmatic approach is required."
   
   "Pragmatic" here means a small, relatively cheap system that's
   flexible enough to be operated by non-specialists. One way of keeping
   things flexible and cheap is to cut out one of the three major
   problems in building an assembly system: control. That's the tack
   being taken by scientists at the University of Tokyo's Research Center
   for Advanced Science and Technology (RCAST) in a system that's
   operated by telecommunications.
   
   Having a human operator gives researchers one big advantage: It allows
   them to concentrate on sensing and actuation rather than on
   programming the robot to do specific tasks. On the other hand, the
   person using the system must be able to see what he or she is doing,
   one way or another.
   
   For vision, therefore, the Japanese system uses two microscopes, one
   optical and the other a scanning electron microscope. Although the
   electron scope requires that the setup be confined to a vacuum
   chamber, it has the advantage of supplying resolution--almost down to
   nanometer dimensions--when required. For its part, the optical
   microscope is needed to orient the user to the work space: Searching
   an area as small as a square centimeter can take a long time when
   using a window only a few microns wide.
   
   Depth perception becomes another issue. In order to allow the user to
   understand the space in three dimensions, the work area can be rotated
   with respect to the microscopes. Showing different angles lets the
   operator pick up parallax information that the eye requires to make
   sense of a 3-D scene.
   
   To increase the realistic feel of macroscopic objects, the human
   controller uses a pencil with force feedback to move objects.
   Supplying the nervous system with scaled-up force information has
   proven vital for dealing with such tiny, delicate objects. As in
   everyday situations, touch is able to fill in when vision is not quite
   adequate. Thus, even if the operator cannot see precisely, it is still
   possible to make contact with the part or determine if the force is
   too great.
   
   This level of interaction is particularly important given the kind of
   applications the RCAST team aims to tackle. Tomomasa Sato, the senior
   researcher in the group, calls microassembly techniques "indispensable
   for realizing real, functional three-dimensional microsystems."
   
                        [LINK] Go To Previous Story
                                      
                          Go To Next Story [LINK]
                                      
                     [INLINE] Back to This Week's News
-- Lloyd Rasmussen
Senior Staff Engineer, Engineering Section
National Library Service for the Blind and Physically Handicapped
Library of Congress 202-707-0535
(work) lras@loc.gov www.loc.gov/nls/
(home) lras@sprynet.com



This archive was generated by hypermail 2b29 : Sun Dec 02 2012 - 01:30:04 PST