The United States has lost much of its industrial power to people who make better refinements of existing products, says IBM's senior vice president for science and technology.

Speaking at Utah State University's 14th annual Productivity Seminar, Ralph Gomory said products are improved, not by some great idea, but by refining. "This is the normal way products compete with one another."The nation's competitive problems are not in the use of new science, according to Gomory, but in incremental improvements.

"When it comes to continuous improvement, we lose," he said.

Gomory was one of several speakers last week at the seminar sponsored by the USU College of Business Parner's Program and the Phi Beta Lambda business fraternity. The focus of the two-day seminar was productivity strategies for the 1990s.

Though the United States is the greatest scientific power in the world, Gomory said he believes industrial power came long before scientific power.

"We were the leading industrial power in the world by 1920, maybe earlier," he said. "However, before World War II we were not a very important scientific power. Japan is now a strong industrial power but not a leading scientific power. You can win in the product competition without being a scientific power."

The dominant people in industrial competition are the product developers, Gomory said. To succeed, he added, businesses must break down the separation between development and manufacturing.

He spoke about the cycle between development and manufacturing and said the speed of the cycle is very important. Products on the faster cycle have the more recent technology.

What makes for a fast cycle? "You have to encourage designs with manufacturability," he said.

He used an IBM printer as an example, saying engineers took apart competitors' printers and found they had fewer parts. He said they designed their printer with fewer parts and with "no springs and no screws" so robots could assemble them.

"The printer became so easy to assemble we almost didn't need the robots," he said, "and that printer became our best seller."

Gomory said product innovations must come from within the development cycle and the engineers developing them have to maintain close contact with manufacturing.

"In the U.S. we tend to separate design and manufacturing," he said.

"It doesn't work. Ideas won't trickle down."