A HUNDRED years ago, the clock tower at the world's largest wool mill helped workers get to their shifts on time.Today, the clock still looms high over these brick buildings 30 miles west of Boston, but work is organized around the quickening cycles of developing new products and bringing them to market. And the business is no longer cloth; it is computers. The buildings, still called "the mill," have been remodeled into a labyrinthine home for Digital Equipment Corporation, one of the world's biggest computer companies. DEC is struggling, like many companies in the industry, to keep up with rapidly evolving technology and markets. For example, the capacity of data storage disks has been increasing at a rate of 20 percent per year. Digital's storage disks have shrunk in size from 9-1/2 inches to 5-1/2 inches, and Charlotte Fredericks is looking toward a 1-1/2 inch size. "I'd like to see evolutions [of the product] every six to nine months," says Ms. Fredericks, DEC's vice president of storage technology. The computer industry is one of the most striking examples of fast-paced innovation. But across the manufacturing sector, companies are learning to develop new products more quickly, and to develop manufacturing systems that can adjust quickly to the marketplace. Seiko, the Japanese watchmaker, "introduces a new model every shift," says Morris Cohen, co-director of the center for manufacturing research at the Wharton School in Philadelphia. Where some companies would employ researchers to predict what people will buy, Seiko can simply take note of which styles sell, and make more of them. The move toward shorter product life-cycles and segmentation of markets is "a fundamental shift that we're going to live with for some time," Mr. Cohen says. "Nobody is expecting to go back to the way things were." Traditionally, manufacturers vied to be the low-cost producer, says Joseph Blackburn, professor of management at Vanderbilt University in Nashville, Tenn. Now, companies still compete on cost and quality "just to be in the game," but they also compete on variety and speed of delivery, says Mr. Blackburn, author of a recent book on "Time-Based Competition." "We need the same 'lean production' system in product development that we have in manufacturing," he says. Lean production on the assembly line refers to the ability to build products in small batches - adding variety without sacrificing the economies of mass production. In product development, Blackburn says the same approach is needed: Process information in small batches. In other words, people from the design, manufacturing, marketing, and field-service functions work together from the outset, rather than passing their completed work "over the wall" to the next group. "Concurrent development" and "cross-functional teams" are two of the current terms used to describe this effort. The end result is a product that not only reaches the market quicker, but also takes into account the concerns of those who will build it, sell it, and service it. In many cases, companies are even getting outside suppliers of components into the act early on. Chrysler and Boeing are two examples among numerous US companies that are developing their own versions of this approach, says Kent Bowen, an engineering professor at the Massachusetts Institute of Technology. Here at Digital, the different functions work together in what Al Cassista calls "virtual teams meaning that team members are linked by computer network rather than working in one location. The product manager can pull together team members from literally anywhere in the world where DEC operates, notes Mr. Cassista, a principal engineer for Digital's new products operations, which makes prototypes. The worldwide computer network came in handy in the mid-1980s, when Digital was trying to become a leader in high-performance storage disks, which expand the data-storage capacity of computers. Keith Glick, engineering manager, says DEC decided to do much of the design work at the manufacturing sites for the key parts: 1) disk platters (the actual storage area) in Tempe, Ariz.; read/write heads in Shrewsbury, Mass.; circuit boards in San German, Puerto Rico; final assembly in Colorado Springs, Colo., and Kaufbeuren, Germany. In addition, "we attempted to involve the manufacturer of each component in the design of the component," Mr. Glick says. The disk, called the RA90, was introduced in June of 1988 and quickly became one of DEC's largest revenue producers. But the process never really ends. DEC introduced upgraded versions of the RA90, replaced it with successor products, and is now working on further shrinking the size of the disks. Failure to keep up with competitors in innovation can mean rapid loss of market share. The question, as Fredericks puts it, is "how do you ride the wave in and not get dumped?" Concurrent development projects require more give and take across functional lines than many companies are used to. After getting used to the sytem, "you spend considerably less money," Fredericks says. Cassista points proudly to DEC's team approach, objecting when a reporter refers to a product manager as a team's "leader." But Blackburn and other experts say product development efforts work best with "heavyweight" project managers. "The alternative is to go slower," Blackburn says. The size of a team, he says, is not as important as who heads it up and what functions are represented. "If you leave marketing out, you're in serious trouble.... They bring the voice of the customer into the process."