Businessweek Archives

An Ivory Tower That Spins Pure Gold

Science & Technology: Research

An Ivory Tower That Spins Pure Gold

As the R&D arm of Lucent Technologies, Bell Labs has a big hand in profits

In February, Tony Tyson was perched in the foothills of the Chilean Andes, trying to solve one of the universe's greatest mysteries. The billions of stars we see represent only one-tenth of the mass of the cosmos. The rest is enigmatic "dark" matter, detectable only by the way it pulls galaxies or bends light. Tyson has pioneered a way to use telescopes, high-tech cameras, and sophisticated imaging algorithms to spot the light-bending, thus charting dark matter's place in the heavens. It's a classic quest for new knowledge--and the stuff of Nobel prizes. But the 59-year old cosmologist is no ivied-hall academic. He's one of 1,150 scientists and technicians at Bell Laboratories, the research arm of telephone-equipment maker Lucent Technologies Inc.

Hold on a minute: How can a company that's in the throes of cutthroat competition afford to contemplate the secrets of the universe? The answer lies in the hallowed laboratories of the Murray Hill (N.J.)-headquartered company. Since AT&T's telephone-gear business and much of Bell Labs were split off to become Lucent Technologies in 1996, the scientists at Bell Labs and the business people at Lucent have forged a distinctive symbiotic relationship. That is helping to fuel the revival of Bell Labs and the surging financial performance of Lucent. "Lucent is the best thing that happened to Bell Labs, and Bell Labs is the best thing that happened to Lucent," says Ravi Sethi, vice-president for computing and mathematical sciences research and chief technical officer at Lucent's communications software business.BUDGET PEG. Credit careful planning with getting the two groups together. At the time of the spin-off, Lucent's then-Chairman Henry B. Schacht underscored the importance of the research effort by putting Lucent's corporate headquarters at the Bell Labs facility. He backed up the gesture with money: He pegged the labs' budget at a fixed 11% of Lucent's revenues. That way, scientists get more research money as the company grows--a strong incentive to help the company's performance. Since Lucent's spin-off, Bell Labs's budget has increased 42%, to $3.7 billion, for the fiscal year ended Sept. 30, 1998. That's when the company posted revenues of $30.1 billion and net income of $970 million.

Lucent also established an internal venture-capital operation to fund researchers' ideas that don't fit into existing business units. Scientists get their ideas financed--and they get equity. One new venture, Visual Insights, sells software that can detect billing fraud by analyzing patterns in large amounts of data. Another, Verdicom, does fingerprint authentication.

The result is that today's Bell Labs is a hotbed of innovation--at a time when Lucent needs leading-edge technology in order to compete in the fierce market for communications gear. "They've succeeded in making an organization that does both basic research and development that's of use to the company," says Stanford University electrical engineer David A.B. Miller. Venky Narayanamurti, professor and dean of engineering at Harvard University, agrees: "Bell Labs seems to have once again found its bearings."

Lucent is the prime beneficiary. Bell Labs research is responsible for 50 Lucent products that are on the market now or that will come to market in the next few months. For example, Tyson's work in charting the heavens helped lead to a fingerprint-identification system. That's because the technology used to chart galaxies can be adapted to analyze other complex images, such as fingerprints.

Lucent and Bell Labs have had some catching up to do. Tiny Ciena Corp. in Linthicum, Md., blindsided the company in 1996 with an optical networking product that doubled the capacity of long-distance phone lines. And Cisco Systems Inc. has built a considerable lead on Lucent in the market for data-networking equipment. Lucent has reacted with the kind of speed it lacked when it was part of AT&T. After Ciena's success, Bell Labs scientists developed a competing product in 15 months. And in January, Lucent agreed to buy Ascend Communications Inc. in order to get data equipment products to compete with Cisco. "Lucent has identified the problem and taken the right steps to refocus its formidable [research] weaponry," says analyst Paul Sagawa of Sanford C. Bernstein.NATIONAL TREASURES. The tale of Bell Labs is more than just the success story of one company. It's also the story of the rejuvenation of America's great industrial labs. IBM's T.J. Watson Research Center, RCA's Sarnoff Research Center, and especially Bell Labs, were once viewed as national treasures blazing a research path for the entire country. Bell Labs, for instance, helped create such breakthroughs as the transistor, the underlying technology of the Information Age, and the optical amplifier, key to today's fiber optic networks (table). Eleven researchers have won the Nobel prize in physics for work done at the lab dating from 1927.

But as competition from Japan and Europe intensified in the 1980s, companies could no longer afford these ivory towers of science. After the 1984 break-up of AT&T, Bell Labs began a period of downsizing and decline. "We really had to understand the difference between long-term research valuable to the company and academic research--and we needed to prune the academic part," says then-research chief and Nobel laureate Arno A. Penzias. That brought howls of outrage from science-policy mavens. "The big industrial labs don't exist anymore," fretted William J. Spencer, then CEO of the semiconductor industry's research consortium, Sematech, in a 1996 interview. "Where are the next breakthroughs going to come from?"

The answer seems to be: from the new Bell Labs. Its researchers are probing the neural pathways of the slug in order to understand how biological lessons could lead to self-healing networks. They are exploring computers that harness the arcane power of quantum mechanics and the information storage capacity of DNA. "Our role is absolutely clear: We are the innovation engine for Lucent," says Arun N. Netravali, Lucent executive vice-president for research.

Researchers have speeded up the steps from lab to market. Consider how Lucent reacted after being humbled by Ciena. The explosive growth of communications demands that more and more information be crammed through every strand of optical fiber. One way to accomplish that is to boost the number of colors of light speeding through the fiber, each carrying information. For years, Bell Labs researchers had been building systems that could carry a handful of wavelengths. Then, Lucent was upstaged by Ciena in 1996, when it sold Sprint Corp. technology capable of handling 16 wavelengths of light--twice as many as the systems Lucent was selling.CROSS-FERTILIZATION. So a Bell Labs team led by Rod Alferness set to work. The central challenge was this: Light gradually fades as it travels through an optical fiber and must be periodically boosted with so-called optical amplifiers. But the company's existing amplifiers couldn't handle more than a few colors. So the team began searching for new materials and approaches to broaden the amplifiers' capability. By mid-1997, they had a breakthrough--a device that could amplify 100 different colors.

Under AT&T, that innovation would have been handed to a development team, taking years to become products, says Harry Bosco, head of optical networks at Lucent. Instead, "we laid the gauntlet down to go for an 80-wavelength network, and do it in a year," he says. The secret was putting people from research, development, and manufacturing together. By early 1998, Lucent was able to announce a 40- or 80-channel system capable of transmitting 5 million voice calls simultaneously through a single fiber.

The flexibility that allows the same employee to wear different hats is a hallmark of the resurgent Bell Labs. For instance, a research team headed by Vijay P. Kumar, head of the high-speed-networks research department, came up with new techniques for directing data through the Internet and other networks. The advances boosted the speed at which data can be transmitted--and allowed Lucent's customers to guarantee certain data speeds to their corporate customers. After developing the technology, Kumar's team temporarily moved to Lucent's business unit to oversee commercial development of the product. The result? The product, called IP PacketStar, made it to market earlier this year. On Mar. 4, Lucent announced the first actual sale, a $28 million deal with British Internet service provider X-Stream Network. British Telecommunications PLC and other companies are testing the equipment.

Researchers also are looking years into the future at discoveries that could transform technology. Take wireless communications. For decades, scientists had assumed there were limits to the amount of information that could be sent over a given radio band. But "the hidden assumption was that radios are expensive and spectrum is cheap," says Richard E. Howard, director of Bell Labs's wireless research. So researchers flipped the premise around: What if radios and signal-processing power were virtually free? The answer may lead to a major advance. Howard's team found that they could send information from many antennas at once across a piece of radio spectrum that would normally carry just one message. "If we have 30 or 40 antennas, we can send 30 to 40 times as much information, with the same bandwidth and power," says Howard.

And just where are all those free radios going to come from? Just stroll down the hall to the lab of Dennis J. Bishop. As part of his physics research, he has built tiny silicon machines, like small hands, that can be used to manipulate the magnetism in tiny specks of matter. After talking with other researchers, Bishop realized that the same technology had a number of practical applications. For example, the micromachines may make it possible to build an entire radio--antennas and all--on a single, inexpensive chip. "All of a sudden, radios would be as ubiquitous as buttons," says Bishop enthusiastically. "You could sew them into kids' clothing, so that you could talk to them."

For scientists inside the labs, the new message is loud and clear. Instead of just pursuing their own curiosity, "now all of us feel a responsibility to look for applications of what we are doing," explains 30-year veteran Tyson. Even on that mountain in Chile, Tony Tyson has one eye on Lucent's bottom line.By John Carey in WashingtonReturn to top

Too Cool for Crisis Management
blog comments powered by Disqus