Recently, I attended an Advanced Manufacturing Research (AMR) conference in Orlando, Fla. Centered around enterprise planning, advanced scheduling and manufacturing execution systems (MES), the conference drew more than 700 attendees, many of whom represented the information technology (IT) contingent of Corporate America. Each of these IT professionals had corporate budgets ranging from $10 million to $100 million. The corporate IT segment is where the action is.

Exactly what is this action? Let's make some predictions. Predicting the future is a dangerous game. There have been many predictions of future technologies; most of which have been proven false over time. Here are some:

"640K ought to be enough for anybody."
— Bill Gates, Mircrosoft, 1981

"There is no reason anyone would want a computer in their home."
— Ken Olsen, Digital Equipment, 1977

"Heavier than air flying machines are impossible."
— Lord Kelvin, anti-scientist, 1895

"I think there is a world market for maybe five computers."
— Thomas Watson, IBM, 1943

The reason these are mentioned is that like forecasting the weather, predicting the future beyond several days is a difficult task. While most futurists would argue that their projection accuracy is about 20 percent, it is clearly better than making no projections at all.


Big Waves

With this in mind, several of the "big waves" I see on our horizon are enterprise resource planning (ERP), MES, advanced planning and scheduling (APS), and the ubiquitous World Wide Web (WWW). All in the IT arena.

Other important examples exist where science is being applied to practical engineering use. Chaos and complexity science and genetic algorithms are two examples of this transfer.

ERP deals with more than just the enterprise. ERP must include vendors, customers and a time scale of several years. ERP gives organizations the ability to monitor and control the food chain. This optimizes assets, works in process, finished goods inventory and market response. ERP is the window for management to manipulate the enterprise at any and every level.

What ERP is to the top level of enterprise management, MES is to the intermediate level. MES concerns itself with the management of the locally owned facilities and with the coordination of centers of productivity, typically within the building. MES organizes the plant floor, tracks the plan, and as a result, improves productivity. ERP and MES, when coupled with real-time control on the factory floor (PLCs etc.), constitute the full communication and computer integration between all aspects of the enterprise.

How can this be accomplished? In the short term, the answer lies with the World Wide Web. Imagine that you are on a Boeing 747 at 30,000 feet. You open your laptop and type in the appropriate access codes, and see before you the company's worldwide assets.

How will these changes come about? From vendor supplied packages. What vendors? Which packages? Momentum in the marketplace determines who wins. This process is demonstrated by the argument between COM/COM+ and CORBA. CORBA has tens of thousands of packages shipped, whereas Microsoft has shipped 150 million packages of COM/COM+. Despite other factors, the numbers are so disproportionate that sheer momentum will determine the market winner, independent of the positive or negative aspects of the technology.

Other changes will come about because there are a growing number of new companies coming on board. These new guys are using genetic algorithms, complexity, chaos and data broadcasting. This will significantly change the world as we know it. Being able to follow the market around and use technology to create wealth leads to a substantial improvement in a company's competitive positioning and its bottom line.

The short-term winners will be:

• Agents (behavioral objects in complex systems)
• Knowledge-based approach
• NT operating system
• Dealing with the customer/vendor in a cooperative rather than adversarial manner

Some additional technologies that will have to be considered over the next year are:

• ADSL (asynchronous digital subscriber line)
• VRML (virtual reality modeling language)
• All of the attendant ERP/MES/PLC connections
• Agents and emergent systems
• Catalytic management
• APS (advanced planning & scheduling)
• Emergent management

Another interesting idea on the horizon is called the virtual factory in which a vendor supplied component of the factory is set up in a small, portable unit resembling a construction trailer. These trailers are then managed over the Web where interaction takes place. Much in the way a modern American company might consider Taiwan as its manufacturing department, small virtual factories inside Taiwan would act as machine stations. They would automatically bid against one another for jobs while delivering components directly into the system. In fact, this is practiced to some extent today. For example, air express companies now build facilities at their large customer locations to serve as their product assembly and shipping departments.

Other short term winners will be JAVA and its counterpart at Microsoft. These winners will force us to take notice of them. Every product, every plant, should be specified as or have products developed for it that are both net and JAVA compatible. This is what will enable ERP and MES. We must learn to deal with systems, not components, with processes and not entrenched technologies.

As we stretch forward into the next five to 10 years, we have to invoke a bit of science. These sciences are:

• Chaos
• Complexity
• Emergent programming
• Theory of constraints
• Genetic algorithms and artificial life
• Predictive modeling
• Catalytic management

One of the promising sciences is artificial life (AL) and its sister, genetic algorithms (GA). Several summers ago, I attended an artificial life conference at MIT. Being lazy, my notes were relatively short. Allow me to present my distilled wisdom to you in bullet form:

• Learning is useless in a stable environment.
• Stop directing change and let it happen.
• Activity, NOT function, is important.
• Formality is wrong.
• Organisms need competition.
• Simple rules are sufficient.

In addition to AL and GA, the disciplines of chaos and complexity indicate much promise. Over the past decade, engineers and executives have come to realize that production systems have become more complex, causing a huge increase in lines of software code. The top-down command control architectures have run their course and attention is now being turned towards a bottom-up approach. This latter approach is an outgrowth of common sense, and the science of complexity. These architectures provide more robust systems, fewer lines of code and performance that more accurately reflects the real world.

Meeting of the Minds: Industry and Science

A sampling of conferences and seminars aimed at fostering the exchange of ideas and practical applications between industry and science: Santa Fe Institute (a renaissance learning process concerned with large-scale and wide complexity issues) BIOS (a for-profit partnership between Ernst & Young and individuals at the Santa Fe Institute) R. Morley, Inc. (sponsor of the "Santa Fe Chaos in Manufacturing Conference" in early April) Flavors Technology (a supplier of system software and hardware to accomplish complexity solutions) NIST (National Institute of Science & Technology) NCMS (National Center for Manufacturing Sciences)

This view, however, requires a paradigm shift of the mind. It is not an easy one to embrace. The benefits of early adoption will, however, yield a competitive advantage that allows for market capture and market share growth. Early lessons from studying ant hills and bird flocks have led to innovative approaches in scheduling, factory control, constraints management and process modeling. These approaches are changing the way software applications are conceived and designed and are providing significant pay back to the companies that deploy them. Several organizations are heavily involved in facilitating the transfer of these sciences into realistic applications with competitive advantages. Many also offer conferences and seminars to foster the exchange of ideas and practical applications between industry and science (see box).


Show-and-Tell

Let's do a show-and-tell: A General Motors plant looking to improve productivity. Why did GM decide to use these new technologies? Because they're better, stronger, faster and yield a competitive advantage through optimal asset utilization. If we can get a software productivity gain of 10 to 1, decrease error rates, increase robustness and better optimize constraints and assets, then why not?

GM is an innovative company that wished to improve its paint facility in Fort Wayne, Ind. Painting is a difficult process that is poorly understood. It is difficult to describe the painting process on a Newtonian basis. It is better understood in terms of complexity science. The solution was achieved by making each paint gun an autonomous agent.

We experienced resistance with the first installation in 1990. Blockers were opposed to non-standardized methods. The burden of proof and the burden of benefits of the installation were difficult to make. Despite this, the first installation achieved an equivalent product fit and finish at 1/12th the number of lines of code. The benefits were direct; more than $1 million in paint was saved annually. GM also won the J. D. Powers Award for the best fit and finish for any truck factory in the world. The system was operator controlled, and fed back directly and immediately. The system was also very robust. The difficulty was that it was not standard.

Why did this solution work? Because we understood the problem and the behavior; we didn't try to decree top-down solutions as to what the plant should be like. Instead, we accepted the plant the way it was and modeled based upon the behavioral aspects. This new approach deals with things the way they are and tries to optimize what is, not what should be. In both of our examples, the system was designed upside down. Scheduling was on a pull-through basis, rather than a push-through. The performance was chaotic — that is, unpredictable but bounded by constraints. How do we actually take constraint management theory and apply it against it a pull-through scheduler to maximize asset utilization? This is what we must address.

We must note that ERP, MES, chaos, complexity, artificial life and genetic algorithms are system considerations and not point technologies. To succeed and excel, we must not fix the peaks of our performance. Rather we should work to improve the valleys. We want a level playing field inside our enterprise. We don't want to be the best at something at the expense of being less than best somewhere else. We must manage not only production constraints, but also technology constraints.


Irrelevant Week

With this in mind, I would suggest that next week is irrelevant. What happens this coming year was predicated upon decisions made several years ago. We have to take actions and make decisions now so that in the next two to five years our enterprises will be improved. It takes 18 months to do any technical turn around system. Independent of negotiation schedules, all engineers know that nothing can be done in under a year. Of course, we can bash out code, make a printed circuit, establish a system. Benefits from dreams, however, take more than a year for any real life creative process. It is difficult to plan for retirement one week before the gold watch.

Returning to our forecasting, let's rattle off some thoughts about the year 2020 (MMXX). Computers will be 100,000 times more powerful than they are now. Unimaginable power in a watch. It means that any algorithm will be on a chip. Behavioral agents will be wide spread. Chip production will come up against quantum effects which will impact their fabrication. On the other hand, we could use quantum theory to make different classes of computers. Bandwidth, because of the 900 satellites now envisioned, will be "infinite." Not really, of course, but it will allow for data broadcasting.

Data broadcasting is the ability to have data transmitted to you at your home or factory; the ability to select the data that goes by for later use. CDs give me a library in my hip pocket. What are the limits? A back-of-the-envelope DNA calculation suggests that we will meet a limit on computing power roughly around the year 2030. I have no idea what technologies will take us there, but I have faith in the human spirit and the technology community.

We've got to remember the slogan, "It's the system stupid!" Towers of excellence violate the theory of constraints in the intellectual flow of knowledge. We must bring all parts of the infrastructure up to the food chain requirements. We must also change how we educate people. Education traditionally centers around small and deep aspects of technology. Chaos, The Santa Fe Institute and BIOS are all centered around broad band infrastructure elements. Their goal is to solve true, real-world complexity issues.

In the short term, we still have to go after cost reduction, short-term scheduling, local constraint optimization and, particularly, time to market. In the long run, we want to make sure we are compatible with all the Web aspects of enterprise management, particularly APS. We have to transform our organization; and transform the process with meta-systems. Total enterprise management is the name of the future. We have to be able to deal with business, engineering, economics, computers and science, all with equal measure. Computers are truly going to change our enterprises, customers, vendors and assets manipulation. Remember, however, technology loses to the market every time. The momentum of the marketplace can't be stopped — only directed. We must learn or die.

About the Author

Richard E. Morley, CEO of Flavors Technology, Inc., and a member of the board of directors of several high-tech companies across the United States, has been a prominent figure in the high-tech industry since the beginning of solid-state electronics. Perhaps best known as the "Father of the Programmable Logic Controller," Morley is a leading visionary in the field of advanced technological development, and an authority on the factory of the future. An engineer, consultant and inventor, Morley holds more than 20 U.S. and foreign patents, including the programmable logic controller. An MIT-based background in physics provided Morley the insight to become an internationally recognized pioneer in the areas of computer design, artificial intelligence, automation and futurism. Richard Morley lives on a farm and works out of his barn in New Hampshire. He has raised more than two dozen children, loves skiing and rides a Harley Davidson Sturgis.

Editorial E-mail:



Web Site © Copyright 2020 by Lionheart Publishing, Inc.
All rights reserved.


Lionheart Publishing, Inc.
2555 Cumberland Parkway, Suite 299, Atlanta, GA 30339 USA
Phone: +44 23 8110 3411 |
E-mail:
Web: www.lionheartpub.com


Web Design by Premier Web Designs
E-mail: [email protected]