It's 1999, and you want to invite some friends to the biggest New Year's Eve party of the millennium. You know they're not home, but you call anyway, and the intelligent network zaps the call across the country to where they're staying. Now, how to describe what you'll be wearing? No need to. This is a video call, so just hold your evening wear up to the camera. As for ordering some fin-de-siecle champagne, have your computer check the Bubbly Data Base, which keeps up-to-the-second data on what's available at thousands of liquor stores across the country.
The phone network of the future will change people's lives in more important ways and will begin to do so well before the turn of the millennium. It will improve health care in rural hospitals by letting big-city medical experts examine CAT scans via video hookups. It will reduce traffic congestion by allowing more people to work at home. And it may make business more efficient--by squeezing out the dead time consumed by letters sitting in the mail and by relaying information continuously to headquarters to avoid inventory pileups. Meanwhile, cellular and satellite phones will link people wherever they are, and vast new capacity will push down prices so far that casual chats will be as routine across the ocean as across town.
In the U. S., Japan, and Western Europe, governments and business interests have come to realize that flexible, high-speed, and reliable phone networks are critical to economic growth. And while those wealthy nations plan for 21st century telecommunications miracles, developing countries from Poland to Indonesia are working toward much smaller miracles--a time when they don't have to wait years to get a telephone line installed or hire secretaries to sit on the telephone all day waiting for a dial tone.
NEW GENERATION. Making these dreams come true will require hundreds of billions of dollars' worth of new phone equipment--everything from picture phones for your den to earth-orbiting satellites. Most critical, however, will be a brand-new generation of phone switches and transmission gear that phone companies from Singapore to Cincinnati will use to modernize their networks. Says Frederick D. Ziegel, an analyst at the New York brokerage of Punk, Ziegel & Knoell: "Over the next 5 to 10 years, private and public switched telephone networks will undergo the most radical technological transformation in history."
The new hardware is needed to complete the broad telecommunications superhighways that customers are demanding. Today's networks are built around the telecom equivalent of an IBM mainframe computer: the central-office switch, a multimillion-dollar monster that's ensconced in a fireproof building and tended to by technicians around the clock. All the phone calls from miles around pass through it.
DIGITAL BLIZZARD. But even these giant machines and the transmission gear they work with can't possibly handle the traffic that's coming down the pike. Soon, the network will be called upon to carry just about anything that can be digitized, from the human voice to high-definition television images. Phone companies are making huge investments to carry those signals by replacing copper wires with fiber-optic cables with theoretically infinite capacity. So, what phone companies need from equipment makers are new switches, transmission gear, and software to keep the deluge of digital signals flowing smoothly over those capacious fibers.
Simply building bigger, faster versions of today's equipment won't do the trick, researchers say. Today's network switches, which route calls from place to place, wastefully reserve a fixed pathway between two points even if the circuit is idle for most of the call. In a typical voice call, the two-way circuit is idle about 60% of the time, figuring in pauses and the fact that usually only one person is speaking at a time. Plus, valuable seconds are wasted in setting up a circuit for what may be a split-second data burst.
The coming generation of switches will break down all messages into tiny packets of digital information that will be "mailed" at light speed over the network and reassembled into a coherent stream at the far end, a streamlined version of the packet-switching technology now used in computer networks. With that change, prices should fall, reliability should increase, and, voila: a superhighway for the Information Age.
Sounds like a bonanza for the makers of phone equipment, right? Not necessarily. While demand for new equipment will be at an all-time high, new technology and intense competition among global competitors are driving down prices, limiting estimated revenue growth to 7% annually through 2000. And profits--already squeezed by price wars in the past few years--will be even harder to come by. To deliver the new technology that the phone companies want, equipment makers are spending huge amounts on research and development (chart). FEW SUBSIDIES. The inevitable result: From around 10 diversified equipment makers now competing internationally, perhaps five or six will survive as global rivals with a full range of products by the turn of the century. As the decade begins, American Telephone & Telegraph Co. and Canada's Northern Telecom Ltd. are the reigning suppliers in North America, the world's most advanced market. In Europe, where national phone companies often still favor local manufacturers, Germany's Siemens, France's Alcatel, and Sweden's Ericsson are well entrenched. And all eyes are on the Japanese, led by NEC Corp. and Fujitsu Ltd., whose deep pockets, persistence, and skill in manufacturing serve them well.
The action is already intense in the U. S., the world's most open telecommunications market, but fierce competition is spreading--even to Japan's Nippon Telegraph & Telephone Corp., which is feeling the prod of competition and is buying more equipment from foreign suppliers to meet the challenge (page 144).
In the emerging global market, the focus of competition is shifting from political influence to technology. While protectionism and strong ties between local phone companies and equipment suppliers linger, few nations can still afford to subsidize equipment makers or live with outdated phone technology.
For businesses, the advent of new generations of superfast switches built on packet technology will mean not having to lease an expensive, high-capacity circuit by the month just to send an occasional flurry of information. Holding open an entire high-speed circuit for irregular bursts of data traffic--as they do now--is like reserving a separate highway lane for a race car that comes through once a month. With fast-packet networks, they'll just dial up and pay for the number of packets sent--the ultimate in what's known as "elastic bandwidth." Inevitably, that flexibility will encourage more high-speed information exchange between people who otherwise would have settled for a phone call or a fax. The first fast-packet services from carriers began appearing last year.
The race to commercialize new technologies such as fast-packet switching could rearrange the competitive balance in the $102 billion-a-year telecommunications equipment market. Who will win? At the start, France's Alcatel leads with equipment revenue of more than $17 billion, followed by AT&T at around $15 billion. After them, in order, come Siemens, NEC, Northern Telecom, Ericsson, Hitachi, and Fujitsu, according to McGraw-Hill Inc.'s Northern Business Information research unit. AT&T, Alcatel, and Northern Telecom have all publicly vowed to be No. 1 at the turn of the century--although Northern hedges that it expects to be the "market leader" in products, not necessarily in revenue.
For any of them to make good on this promise will require something that none yet has: a mastery of global marketing. Until the 1980s, they didn't need to think globally. Government-owned national phone companies bought mainly or exclusively from a handful of favored domestic suppliers. AT&T, barred for decades by antitrust decree from selling abroad, sold nearly the entire output of its Western Electric Co. to its own long-distance and local phone operations.
INVADERS. The 1984 breakup of AT&T stimulated the phone giant to go after foreign markets, and other manufacturers followed suit as they took advantage of the 1980s' wave of telecom deregulation and trade liberalization. Western Europe, long sealed off from competition, is cracking open as monopoly phone companies begin to face competition and the European Community works to create a single borderless market.
That already has sent a tremor of consolidation through the industry. On the theory that the best defense is a good offense, suppliers invaded each other's markets while staving off assaults on their own. Siemens has bought the Rolm PBX manufacturing business from IBM and controls British equipment maker GPT through a partnership with General Electric Co. of Britain. AT&T is slowly digesting GTE Corp.'s U. S. switching business. And last year, Northern Telecom spent $2.8 billion to complete its buy of STC PLC, a British equipment maker.
No. 1 Alcatel has been particularly hungry lately, taking control of the transmission-equipment businesses of both Rockwell International Corp. and Italy's Telettra. Now, according to Northern Business Information, Alcatel is the leading candidate to take over DSC Communications Corp., the Plano (Tex.)-based switchmaker that has faltered since its signaling software was blamed for causing major phone outages last June in Washington, Pittsburgh, and Los Angeles.
UNCERTAIN PAYOFF. Clearly, all the big players can't be No. 1. Equally obvious, not everybody in the telecom equipment business now is going to make it to the 21st century. The biggest hurdle could be simply funding the huge R&D efforts needed to perfect the switches for the new phone networks (chart). The cost of developing a new generation of switches is at least $1 billion. Says Robert W. Lucky, AT&T Bell Laboratories' executive director for communication sciences research: "It's eating us up. We're pushing close to the frontiers of nature, when you're working with atoms and picoseconds and gigabits. You try to touch one of those gizmos, it costs you an arm and a leg." And costs are even more staggering in software, which now accounts for 75% of the switchmakers' R&D budgets.
What's more, to survive, most big network equipment makers will have to expand into full-line suppliers. As transmission gear gets more capable, it interacts in complex ways with switches, so companies need to tie together the entire chain of equipment, including a network-management system that lets operators survey and fix problems over the entire system. "This is the next wave," says Jozef Cornu, executive vice-president for Alcatel. "It's pure software, and only the very large groups can possibly afford to develop it."
All that big spending, however, may not see a ready payoff. Phone companies, which were once all-but-captive customers, are learning to shop around for the best price, leading to intense price wars in some markets. In the U. S., according to Northern Business Information, the price of a central-office switch has fallen more than 20% in the last two years, to about $220 per line. While the research firm expects prices to level off, an executive for one major manufacturer frets: "In 18 months, prices are going to be in a free-fall."
Competitors blame AT&T for the price war. Says E. Van Cullens, senior vice-president for marketing and business development at Siemens Stromberg-Carlson: "They seem to have taken an attitude in the last year that they will price at whatever level it takes to hold or gain market share." While declining to respond to that, AT&T Network Systems Group Executive William B. Marx Jr. concedes: "Certainly, we've gotten tougher. We'll do what's necessary in terms of pricing."
Against a backdrop of plunging prices and escalating costs, even the best-financed phone-equipment makers must plot careful marketing strategies. In phone equipment, as in computers, major technological shifts create an opportunity for upstarts to grab market share from established leaders. Northern Telecom, for example, cracked AT&T's monopoly on the U. S. switch market in the 1980s by beating it to the punch with digital switches. The lesson: Even top suppliers can't afford to be late in adopting new technology.
But being first can be dangerous, too. Two years ago, Fujitsu was the first to trot out a new generation of fiber-optic transmission products, known as Sonet, for Synchronous Optical Network. The Japanese electronics giant--followed by Alcatel, Northern Telecom, and others--began selling Sonet gear. But AT&T held back. It missed some early orders, but AT&T executives maintain that they'll come out ahead because they may wind up being first with hardware that fits a new Sonet standard. It's not yet clear whose strategy was smarter.
Fujitsu may reap more by bolting ahead with a new fast-packet switching technology known as Asynchronous Transfer Mode. ATM, which will start appearing in phone networks around 1993, breaks all traffic into tiny, evenly sized packets, known as cells, for maximum routing speed. Fujitsu's basic switching design for ATM is primitive compared with those of American researchers, but it makes up for that by using superfast integrated circuits, says Polytechnic University of New York professor Tony Lee, a former researcher at Bellcore, the Baby Bells' R&D arm. "It seems that they don't care what the perfect solution is; they will go ahead to build first," says Lee. "This is a very good strategy because you want to occupy the market earlier." Fujitsu is already conducting trials of its ATM technology with three of the seven Baby Bells: BellSouth, Nynex, and U. S. West.
LONG DELAYS. The accelerated rate of change in the switch market is making it look much more like the computer market, with which it shares key technology such as microprocessors. Much as IBM has had to rethink the role of its mainframes in an era of powerful personal computers, switchmakers and suppliers of other phone equipment may have to change their ways of thinking in the 1990s.
Like mainframes, central-office switches are not so easily adapted to a changing market. Switches remain so difficult to program that new features take years to deliver. Take U. S. West, a regional Bell that wants to offer a feature for people who call and get a busy signal: Press 1 to stay on the line, 2 to leave a voice-mail message, or 3 to cut through in an emergency. Sounds simple, but U. S. West estimated that it would take three years or more to offer that simple service--if it relied on its switchmakers to write and test the required software. The programs inside switches are so byzantine that it takes 1,000 programmers to make sure that one new feature won't interact with another in some unexpected way and bring down the entire network, the company says.
So the regional Bell is doing something that all switchmakers are watching carefully. It is going to write the programs for the new features itself. And, striking at the heart of the switchmakers' business, it expects them to run not on the phone switch but on a Hewlett-Packard Co. computer. If the experiment works, it should vastly simplify testing how new features will interact with existing ones, cutting the waiting time for new services to three to five months and the number of programmers required to around a dozen, says Lloyd Linnell, U. S. West's program director for network delivery capabilities.
It may seem like a small event, but the potential fallout is enormous. If U. S. West succeeds, over time other phone companies also can start shifting some of the intelligence needed to run their networks from central switches costing more than $1 million to minicomputers costing as little as $100,000. Companies such as HP and Digital Equipment Corp. would vie for the new business, while switchmakers could see their world shrinking, perhaps reduced to handling the mundane tasks of routing calls to and fro. Switchmakers already acknowledge that opening the network to other equipment is inevitable. In fact, AT&T is going with the flow and supplying another regional Bell, BellSouth, with a system similar to the U. S. West setup that puts new network features on processors separate from the switch.
U. S. BATTLEGROUND. For the established switchmakers, another alarming aspect of the U. S. West project is the company it chose to furnish the basic software: NEC Corp., one of those Japanese companies that supposedly can't write software. Although the work was done by a team of programmers at NEC's Irving (Tex.) plant, NEC will now be able to sell such software wherever it sells telecom equipment.
The U. S. market is where the great international phone-equipment contest is getting its start. Not just NEC and Fujitsu, but Siemens, Alcatel, and Ericsson continue to push hard into the highly deregulated U. S. market--despite vicious price competition and single-digit sales growth. The reason? "America has the most advanced telecom services in the world," says Ryoichi Sugioka, president of Fujitsu Network Switching of America Inc. "Fujitsu's customers in Hong Kong and Singapore want those services, too. Our experience in America will show Fujitsu in a good light."
Of course, U. S. suppliers don't plan to be sitting ducks. AT&T is accelerating an overseas push that began in 1980 after a 55-year absence from foreign markets. The goal is to get half the company's revenue from overseas calling and foreign equipment sales by early in the next decade, up from 22% now. So far, its biggest telecom-gear markets are in such nations as the Netherlands, Spain, and South Korea, but it has recently made sales to NTT in Japan, too.
Other U. S. companies are also racking up respectable international sales, especially in the new wireless technologies. Motorola Inc. is waging a worldwide battle with Sweden's Ericsson for leadership in sales of cellular network gear. The U. S. company already leads in sales of the phones themselves. In satellite communications, U. S. companies, led by Hughes Communications, dominate the world market.
TOUGH NUT. For U. S. suppliers that are selling switches and transmission gear for the wired phone network, however, Europe remains a difficult market. Most telecommunications gear is still purchased by giant, government-owned phone companies that continue to have close ties to favored suppliers, preferably domestic. Americans claim that the high prices Europeans charge in their home markets enable them to subsidize attacks on foreign markets. AT&T Chairman Robert E. Allen last year cited a news report that Siemens charged $700 per switch line in Germany and less than $250 a line in the U. S. Siemens says that's not a fair comparison because the German price includes software.
The Japanese have had a particularly hard time cracking the European market, and it's not clear how quickly the barriers will fall after 1992. In Germany, Siemens is lobbying to exclude from the European Community any products that don't have at least 50% local content. It is arguing that high-wage Europe will be flooded with imports from low-wage Asian nations--even though Japan clearly isn't a low-wage nation. Says Siemens Executive Vice-President Hans Baur: "The question is employment in Europe. It's a political question. Europe will lose jobs."
That leaves developing countries, mainly in Eastern Europe and Asia. There's the lure of booming growth rates in certain nations, but it takes Herculean efforts to land contracts. Today, the Japanese are consolidating their influence in many of the developing Asian countries through investment, patient market development, and financial aid. Alcatel and Siemens are vying for leadership in Eastern Europe by offering jobs and a transfer of technology through local joint ventures. Companies such as AT&T and Northern Telecom are scrapping for sales wherever they can land them.
In short, safe markets have become unsafe, and well-understood technologies are becoming obsolete. But then, that's what happens when a market pushes the edges of sophistication. Indeed, companies that get it right can still expect to profit. Says AT&T's William Marx Jr.: "The `90s hold tremendous opportunity for businesses like ours. We're not in any way, shape, or form intimidated by the fact that it's competitive." That's what they all say. Some will be right.Peter Coy in New York, with Jonathan B. Levine in Paris, Neil Gross in Tokyo, and Gail E. Schares in Bonn