Intelligent Manufacturing April 1995 Vol. 1
No. 4
The most important technology for manufacturers in the upcoming years
is software, according to a Delphi study conducted by the National
Center for Manufacturing Sciences (Ann Arbor, Mich.). More
significantly, the types of software manufacturers will be most
concerned with are those we think of as "intelligent," i.e., neural
networks, autonomous agents, expert systems, chaos theory,
object-oriented, etc.
According to Dick Morley, president of Flavors Technology
(Manchester, N.H.) and the keynote speaker at the Advanced
Manufacturing Technology Conference held last month in Cleveland,
Ohio, "Only one thing creates wealth - technology and the ability to
use it. In ten years, computers will be 100 times more powerful than
they are now, at one/tenth of the cost."
Morley, a well-traveled technology futurist, expounded on the
Japanese view of manufacturing in the first decade of the next
century:
"Even here in the U.S., people are talking seriously now about
automobiles being manufactured right at the dealership," Morley said,
"in a room no larger than a typical meeting room."
Fanciful predictions are part and parcel of a futurist's resume, and
yet Morley's remarks were grounded in the real world. Speaking to an
audience made up of manufacturing personnel, Morley reminded them,
"This year's profit is based on decisions you made five to seven
years ago." Whether we're able or willing to accept that ours is the
Age of Information is beside the point. "Embrace the technology and
embrace the future," Morley urged, "because that's where you're going
to live."
Shifting Paradigms
Technology can be a ruthlessly ambiguous concept, however. In a very
real sense, for many manufacturers technology is just another way of
referring to what your competitor has, but you don't. The question on
the minds of the conference attendees was, "Why should we invest in a
new technology when we're just barely able to manage day-to-day
operations as it is now?"
Attempting to answer that, James Heaton, president of AMR Consulting
( Boston, Mass.), took a matter-of-fact approach to the situation.
"Paradigm shifts only result when there is sufficient dissatisfaction
with existing technologies, and there is an acceptable alternative,"
he said.
Heaton's view of computer technologies focuses on three application
paradigms:
"The paradigm shift from Phase One to Phase Two is so dramatic,"
Heaton warned, "that if your competitor gets a year and a half ahead
of you, you may never be able to catch up." However, he continued,
the amount of time and money a manufacturer has invested into
existing technology inhibits them from pursuing newer technologies.
Once a company reaches a certain size - becomes respected and
successful, in other words - its very size becomes an anchor weighing
down its chances of becoming even more successful.
Addressing this paradox, Mark Roth, president of Framework
Technologies (Wellesley, Mass.), emphasized that the important point
is not the technology - "It's how are we going to make it work. The
problem is adoption."
Continuing this theme, Roth stated, "We need to design systems that
accept the new and emerging technologies and then improve upon them
by adapting them to existing manufacturing requirements." He
predicted that manufacturing system architectures will need to be
decentralized, asynchronous, agent-based, and self-organized.
Such systems include the bleeding-edge technology known in some
quarters as holonics (see sidebar), which is an attempt to craft
software that, like natural creatures, learns from its mistakes and
past behavior to improve its performance over time. Holonics embraces
the idea that a complex system that works has to begin with a simple
system that works.
According to Odo Struger, vice president of technology at
Allen-Bradley (Milwaukee, Wis.), the technological growth worldwide
is moving toward holonic systems. Struger sees some of the major
technological challenges for the early part of the 21st Century to
be:
Extrapolating from the current rate of growth in computing speed
and performance, Struger predicted that by the year 2020, computers
could be as powerful as the human brain. By the year 2040, computers
could begin to reach beyond the capacity of the human brain. What
will be lagging behind, however, is the software.
What Does It All Mean?
So what does the advent of technological change mean for
manufacturers? Besides the inevitable pitches from consultants and
vendors to invest in their particular vision of the future, a few
pearls of wisdom came from David Goselin, vice president of
operations at AEG Schneider Automation (North Andover, Mass.). To be
effective in any technology shift, Goselin said, "You have to have a
vision, and it has to be shared company-wide. To be effective, you
must qualify the process, not just the product."
In a similar vein, Ernie Vahala, retired group director for General
Motors and the current president of the Auto Body Consortium, pointed
out, "Here in the U.S., we're great innovators, but lousy
implementers when it comes to technology. For manufacturers to accept
any kind of change, they have to see evidence of its benefits in
operation."
There are plenty of vendors out there, very ready and willing to urge
you on to the next level of technology. The most important lesson is:
Make sure you're ready for that next level, and make sure it's the
right level for your company's needs.
What is a holon?
According to Odo Struger, vice president of technology at
Allen-Bradley (Milwaukee, Wis.) and a member of the Holonic
Manufacturing Systems Consortium, a holon is an autonomous,
cooperating agent. The word itself derives from a description of
cellular automata in The Ghost in the Machine (1989, Arkana Books) by
Arthur Koestler, coined as a compound of "hol" from "holistic"
(meaning "whole") and "on" (meaning "particle") from "proton" or
"neutron."
What is a holonic manufacturing system?
According to Max Winkler of the University of Hannover (Germany),
holonic manufacturing systems are considered as a promising answer to
many requirements related to production and manufacturing. The
ultimate aim is the development of an architecture for highly
decentralized manufacturing systems, built from a modular mix of
(semi-)standardized, autonomous, cooperative and intelligent
elements.
What is the HMS Consortium?
From the HMS WorldWideWeb page (http://www.gse.rmit.edu.au),
maintained by David Chia at RMIT University (Australia), comes the
following information:
The Holonic Manufacturing System (HMS) Consortium Technical Plan
addresses research, pre-competitive development, systematization, and
standardization of architecture and technology for open, distributed,
intelligent, autonomous, cooperating ("holonic") systems for
application in intelligent manufacturing systems encompassing
discrete, continuous, and/or batch processing. The consortium's
principal goal is to establish technological and organizational bases
for the development, implementation, world-wide deployment and
support of holonic manufacturing systems.
Who are the partners in the HMS projects?