What do Ethernet, PCs, and Micro-soft products have in common?

A. They are all commercial computing standards.

B. They have changed the way companies manage business data.

C. They are changing the way manufacturers manage plant data.

D. All of the above.

The answer, of course, is D. The past five years have seen enormous changes in automation architectures—changes that would have been unimaginable without the application of Ethernet, PCs, and Microsoft operating systems in industrial control. No longer is it a question of, "Are these products suitable for critical manufacturing applications?" The questions now asked are, "Have we reached critical mass?" and "What will be the next technology evolution?"

To answer these questions, let's examine three recent observations by industry experts and colleagues:

"94% of all industrial communication is Ethernet TCP/IP and 6% is everything else. Five years ago, the numbers were flipped. All other networks are shrinking. Communications through any other network is a waste of time."

These remarks were made by Dick Morley, ceo of Flavors Technology, at a recent "Strategic Manufacturing and Automation Conference" hosted by AMR Research. What makes Mr. Morley's comments especially interesting is that he is credited with developing that most proprietary of control engines—the programmable controller. Back in the 1960s, Mr. Morley's invention was the only "PC" in automation.

If the majority of industrial communications are via Ethernet TCP/IP, what are all the bus wars about? The next time a vendor pays you a sales visit, ask for an explanation of its networking strategy and what makes it superior to other networking solutions.

The second observation comes from James Heaton, vp of technology research at AMR Research. In his conference remarks, Mr. Heaton suggests commercial technologies will continue to push into the controls market:

"Today's college freshmen have never known a world without an inexpensive desktop PC. We're going to have to retrain them if we want them to use something different than PC technology in the manufacturing setting."

The IBM PC was invented in 1981, just as today's college freshmen were turning one year of age. Some used PCs as early as kindergarten, all used them extensively through high school. The students that enter the workforce as operators, technicians, and engineers know the PC platform. In a labor market characterized by low unemployment and skills shortages, don't even think about installing something other than the PC for plant automation. Retraining is not an option.

The final observation I'd like to share is from a colleague in Internet publishing:

"The good thing about monopolies is that they only last 10 to 15 years. After that, they tend to get so large they disintegrate. It's only a matter of time before Microsoft suffers

this fate."

My, how we love to hate Microsoft. Despite the U.S. Justice Dept.'s investigation, Microsoft shows little sign of disintegrating.

Its influence in the automation market can be summed up in one word—Windows. Without even trying, Microsoft has dominated industrial software development with its Windows 3.x and NT OSs.

With the new WinCE OS and soon-to-be-announced DNA (Distributed iNternet Architecture) for manufacturing, Microsoft has finally recognized the real-time control market. If Microsoft, the monopoly, is eventually broken up, it's legacy in the controls and automation industry will have been long established. As an analogy, consider that we now have competition in the railroad industry, but monopolies ensured that all trains were running on the same gauge tracks.