The Office of the Future is the computer industry's version of Shangri La, a place where managers would harmoniously interact with their high-tech workstations to produce the most profitable of decisions and instantaneously transmit them to precisely the right spot in their companies.

Like Shangri La, however, the OOF, as it's affectionately called, is a mythical ideal. But it's a vision so compelling that hundreds of companies are betting their futures that they can build it. Those that can stand to win the lion's share of a multi-billion dollar market.

Getting to OOF is a complex balance between technology, software and cost. Several companies, such as Wang and Xerox Corp., propose wiring various computer devices in an office into "local area networks."

However, IBM's recent move to both buy a major share of Rolm Corp., a telecommunications equipment firm, and team up with it, indicates that it believes that its quickest path to the Office of the Future may be through existing phone lines.

Rolm is a leading manufacturer of PBXs (for Private Branch Exchange)--the switching systems that traditionally have routed telephone calls within a building. However, says Richard Moley, Rolm's vice president of marketing, "A PBX is no longer a PBX--it's more like a computer."

"The intelligent PBX is really going to be the office of the future," says Robert E. LaBlanc, a telecommunications venture capitalist. "Local area networks are great ideas, but you have to pay for them. You can do almost anything you want through a telephone's twisted copper pair."

In essence, says LaBlanc, a company can build a computer network around the same network it uses for voice communications by using an Intelligent PBX. "Two for the price of one," says Rolm's Moley.

The catch, says Paul O. Lindfors, IBM's manager of systems strategy for telecommunications, is that "The PBX is designed primarily around voice requirements."

The demands of data transmission, he says, are radically different than those of voice. For one, computers communicate in a digital dialog of binary information whereas voice communication relies on sound.

More important, the fundamental nature of data communications is different. Not only do computers communicate at rates far faster than voice, but dozens of them can be in conference at once.

"Think of a computer assembling a document from a number of different files," says Lindfors, "that requires high-speed switching and multiple concurrent connections."

To get around these problems, Rolm and other PBX companies have designed digital switches--switches built for computer communications. The problem there is, though, that voice is not easily transformed into digital impulses. "Digital voice is still something of a black art," said Peter Anderson, president of Ztel, Inc., a two-year old Intelligent PBX company.

Technical problems aside, LaBlanc believes that the breakup of AT&T Co. and the deregulation of the telecommunications industry will drive corporations into the arms of Intelligent PBX companies.

Mitel, which was working with IBM prior to the Rolm announcement, and Northern Telecom are also major Intelligent PBX companies. AT&T, through its American Bell subsidiary, is expected to be a major player. The PBX business is expected to grow at a 25 percent compounded rate through at least the rest of the decade.

However, says Fred Withington, an office systems analyst for A.D. Little. "The evolution of customer demand has been a lot slower than many expected. By and large, in most companies, data and voice are still done separately."

But Withington, too, believes that voice and data communications will eventually be integrated by the Intelligent PBX. He anticipates that the demand for PBXs will accelerate with the growth of computerized data banks that can be tapped by customers over telephone lines.

However, most analysts say that the development of an intelligent PBX won't mean the death of local area computer networks. "Many companies will have both an Intelligent PBX and local area networks," says IBM's Lindfors.