The Idea Factory: Bell Labs and the Great Age of American Innovation

by Jon Gertner

  What also made cellular possible were the phone network’s new electronic switching stations, or ESSs. In 1964, when Bell Labs had opened its first ESS center in Succasunna, New Jersey, the public relations department had urged Bell engineers to explain that the ESS had the ability “to provide services we haven’t even thought of yet.” Six years later, Frenkiel and Engel and the rest of the Bell Labs cellular team were envisioning what those services might be. A cellular phone would need to send a digital signal every few seconds to the nearest base station antenna. The base station would in turn send that information to a mobile switching center. But really that was just the start. A vast amount of data needed to go back and forth, almost constantly, between the mobile phones, the base stations, and the mobile switching center. The switching system would have to communicate with the base stations so as to keep track of who was where. In addition, Frenkiel explains, “every few seconds, signal strength data would have to be taken at surrounding base stations to find out if a better one now existed.” If another base station had a stronger signal, the computerized system would hand the call off to the next cell. The catch was that before a call could be handed off, the switching system had to identify a channel in the new cell, set up that call, and send a message to the mobile so it could switch frequencies. As Frenkiel remarks, the ESS was invented to be a central office switch, meaning it was created to simply direct calls between landline phones.8 “But it’s sitting there and it’s got the ability to be programmed for something no one ever expected it to do—all those instantaneous decisions that were never necessary. And now we come along to say we need to do locating and handoffs. And also, by the way, we need to keep track of the health of every base station in the system, we need trouble reports, we need to gather data for traffic.” The ESS—a switching system with a powerful computer embedded within it—could do all these things.

  Frenkiel, Engel, and Porter could identify these challenges, but they couldn’t solve them. As systems engineers, they were looking at a big project in a comprehensive but somewhat general way. Systems engineers consider all the standards and technologies and economics necessary to make a project work. They worry, moreover, about how to integrate a complex new technology with the rest of the system. Cellular phones were an ideal case, since the new technology had to (1) work, (2) work affordably, and (3) work seamlessly with the rest of the existing phone network. What the systems people couldn’t do was actually make those projects function the way they dreamed.

  Help came from several different directions. Bill Jakes, the lead engineer on John Pierce’s Echo experiment, had already been asked by Pierce to look deeper into the science of microwaves.9 He was in the research department; he was therefore looking for new knowledge. When Jakes set out to work on the cellular research, he bought a van with Bell Labs money and hired a driver. Then he and some colleagues piled in the back with recording equipment and headphones, and the van drove thousands of miles along the highways and byways of New York and New Jersey. “We went everywhere,” Jakes recalls, “through tunnels, up on top of hills, in woods, near lakes. All day long, day after day after day.” Their goal, in part, was to study the effect of obstructions on transmission and reception. Over the course of months, they puzzled over why microwaves behaved one way in a particular situation, and another way in a different situation. The radio transmission problems involving a moving vehicle and a cellular system, Jakes later wrote, were “so difficult they challenge the imagination.”10

  THE WORLD IS FULL OF NOISE. It blocks our attempts to talk with one another. Claude Shannon had philosophized about this, but to radio engineers in the field, noise is a slightly different phenomenon. You think about noise not so much as an idea that interrupts a message containing information. You think about noise as clicks and static and fadeouts—a physical or electrical problem that must be overcome by engineering or by savvy. Urban noise is often more intrusive than rural noise. Car engines and trains and legions of electronic devices create bursts and blizzards of interference. Also, there are big buildings, throwing shadows over reception areas or bouncing signals unpredictably to and fro. Even in the countryside, though, noise can interfere with radio transmissions. There are highway overpasses and high-tension wires; there are mountains and trees.

  “The guys who made cellular real,” as Dick Frenkiel says, were recruited from the Bell Labs Whippany office, where most of the work was military-related. These men were from the development department, the place at Bell Labs that actually made ideas and new knowledge into innovations. For the cellular project, their job was to build the components—low-power antennas, experimental phone sets, and so forth—that the systems team had described. They would spend most of the 1970s worrying about noise and interference of one type or another. A small but typical concern: How high do you actually need to place an antenna to make it work? One of the early engineers in the cellular development group was named Gerry DiPiazza. He was not the sort of Ivy League type that clogged the hallways of the Murray Hill research labs. Rather, he was an engineer’s engineer, who had gone to a local New York City school and had attended “Kelly College,” the Labs’ continuing education program. He had essentially lifted himself up through the Bell Labs ranks by his bootstraps.

  By 1968, DiPiazza’s career with the Labs had already involved a surreal tour of duty in the South Pacific. In the early 1960s, Bell Labs had established a tiny laboratory outpost, totaling about forty people, on a remote coral atoll, only about 820 acres in size, known as Kwajalein.11 It was part of the Marshall Islands and had been the site of a significant battle during World War II. A largely secret installation, where engineers now worked on radar communications and antiballistic systems known as Nike-Zeus missiles, Kwajalein, or “Kwaj,” as the Bell Labs employees called it, was several thousand miles southwest of Hawaii. In 1965, a Bell Labs vice president had recommended that DiPiazza go there as a good career move, and DiPiazza agreed. He and his wife packed their belongings and their young children. They had never ventured far from New Jersey. They boarded a jet for San Francisco, then a propeller plane to Hawaii, then another propeller plane to Kwaj. They landed on a small airfield in the tropics amid stifling humidity. “We were greeted by a family with a truck, and we were driven to a household,” DiPiazza recalls. “We were given dinner. And then we were taken to a concrete duplex home. All the furniture was rattan. It was unkempt and hot. It needed to be rehabbed. There were rat droppings in the crib. My wife cried all night, and I did, too. She looked at me and asked, ‘What did you get us into?’ ”12 His family was allowed to fly home for a two-week vacation, but only after DiPiazza had worked on Kwajalein for 510 days.

  Kwaj was not only isolated. It was strange. A local Marshallese population traveled to the island each morning on U.S. military launches to work as merchants and maids. The island’s king and landlord carried around with him his rent money—purportedly a million dollars, paid by the U.S. government—in a bag. Bones of Japanese soldiers, now dead for at least twenty years, turned up regularly on the beaches. On the main street of town was a grocery store where everything was sold frozen, even the milk and meat. There were few cars; instead, DiPiazza and his wife rode their bikes everywhere. And in time, they grew to like it. A tight camaraderie existed among the families there, and the leisure pursuits—sailing, swimming, and free movies—helped pass the time. Most important, DiPiazza’s work on highly sophisticated radar systems held his interest.13 Bell Labs and Western Electric were testing the capabilities of the radar systems against target-practice missiles shot toward Kwajalein from Vandenberg Air Force Base in California. At the time, military strategists were concerned that foreign missiles would be built to scatter a cloud of foil decoys—they called it chaff—before detonating. The foil confused radar defenses and obscured the warhead inside the chaff cloud. DiPiazza’s job was to develop techniques that allowed U.S. radar to discriminate between the decoys and the actual explosive, so that a U.S. missile could take out the ene
my warhead.

  The work put DiPiazza on the vanguard of radio engineering. And two years later, when he finished his tour of duty on Kwaj and returned to Bell Labs’ Whippany office, he was asked to continue with military radar. Not long after, though, Bell Labs decided to get out of missile work when it became apparent that the United States would sign an anti–ballistic missile treaty with the Soviet Union. “Bell Labs wasn’t going to fire us,” recalls DiPiazza. “They were going to tell us to find a job within Bell Labs.” Around that time, he ran into a good friend of his at the Labs. “My friend said, ‘You know, you guys should be doing radio telephone.’ ” And so DiPiazza and his boss asked around and soon found themselves in a meeting with Dick Frenkiel and Phil Porter in Holmdel, listening to them talk about cellular telephony. They pitched their own talents in the hope of being brought in.

  There may have been only a few people in the world who knew as much about building high-frequency military communications systems as DiPiazza. And none of them were drafted, serendipitously, into a cellular telephone project. As DiPiazza saw it, the systems engineers like Engel and Frenkiel were standing around blackboards or sitting at their desks with sharpened pencils. His team of development engineers could actually build what they needed—a complex mobile radio that could automatically change frequency, for instance, as a driver moved from cell to cell. For the next few years, DiPiazza, like Bill Jakes, spent most of his time in the field, driving around New Jersey and Philadelphia with his colleagues in an effort to construct a real-world cellular experiment. The men had purchased a small trailer home, then they ripped out the bathrooms and kitchens and hauled in small computers and electronic detection equipment. They called the trailer their mobile technology unit. Often the team would stay up all night, driving through various Philadelphia neighborhoods, testing signal strengths in their trailer home and tinkering with their hardware.

  Mainly, they needed to know what radio signals did in urban and suburban settings. As DiPiazza says, “You had to find out, What is the noise level in a suburban environment? How far would a signal go if the antenna was at ten feet, twenty feet, fifty feet? Would it go one mile, two miles, four miles? How many antennas do you need? How do you build an antenna? What are you going to put the antenna on?” Nobody had ever built those things before. And if a cellular system was going to someday expand across the country, the core challenge was coming up with standards. They had little time to contemplate their legacy, but some of the decisions made by DiPiazza’s team proved indelible—for instance, the height of a cellular tower and the close arrangement of its antennas. Meanwhile, Phil Porter, who had worked with Frenkiel on the original system, came up with a permanent answer to an interesting question. Should a cellular phone have a dial tone? Porter made a radical suggestion that it shouldn’t. A caller should dial a number and then push “send.” That way, the mobile caller would be less rushed; also, the call would be connected for a shorter time, thus putting less strain on the network. That this idea—dial, then send—would later prove crucial to texting technology was not even considered.

  In December 1971, AT&T submitted a long and detailed cellular proposal to the FCC, and the regulators began their deliberations. Motorola, whose radio business could conceivably be jeopardized by the Bell Labs plan, made a variety of seemingly contradictory arguments against it. The company objected to the proposed cellular system on technical grounds, indicating that they believed it simply wouldn’t work; at the same time, Motorola claimed that AT&T would enter the cellular market and use its monopoly power to crush any competition.14 Not long after AT&T submitted its proposal, the company announced that it would only seek permission to build and operate cellular networks. “The company felt it had to make some concession, so they said they would not make handsets,” Dick Frenkiel recalls. Thus the handset business would be opened up to companies like Motorola or Japanese vendors. Such moves were meant to appease regulators concerned about competition. As the phone company executives now realized, the pervasive feeling in Washington was that the current monopoly was big enough. Perhaps it was too much. The cellular business, if it caught on, was going to be different.

  IN 1975, the Bell Labs optical team decided to test their new product. Executives at AT&T and Bell Labs had concluded that fiber, at least at first, would be most useful for high-traffic areas within cities. Even without the Picturephone, copper lines were becoming congested with the steady increase in phone calls and computer data.15 The first fiber was fabricated at a new manufacturing facility in Atlanta. But Bell Labs engineers decided the Atlanta factory would also serve as an experiment—the first building where fiber would be installed.16 The Labs engineering team planned to string a glass cable—actually, 144 fibers arranged on a flat ribbon, and then bound within a protective cover—through the factory’s underground ducts. Afterward, they would spend several months testing the fiber, lasers, repeaters, splicing, transmission quality, and a host of other technical matters. Though it was only the thickness of a person’s thumb, the cable could carry forty-six thousand two-way conversations.

  It went even more smoothly than anticipated.17 Encouraged by the success, the team began a more practical test a year later in Chicago, where fiber would be installed in the field to transmit voice, video, and data to customers and between switching offices. Chicago, like Atlanta, proceeded without a serious hitch. This was cause for elation within a company that was now engaged in a legal struggle for survival. “I have taken very seriously the principles of innovation that you and I followed so long, and that you emphasized about lightguide systems years ago,” Bill Baker wrote at the time to John Pierce in California. Pierce should take pride in the Chicago effort, Baker suggested. “Your spirit is embodied in it.”18

  Chicago was also chosen as the test site for the new cell phone technology. The FCC seemed to be moving cautiously toward approval. But it would first review how the Bell Labs system worked in the real world. In the summer of 1978, Bell set up a working cellular service in Chicago, limited to two thousand paid subscribers, over a two-thousand-square-mile area of the region. For the most part, the test was meant to satisfy the FCC’s desire to better understand the market—not only how cellular subscribers judged the service, but how equipment from other manufacturers, like handsets built by Motorola, meshed with Bell’s new network.19

  The Chicago tests of fiber and cellular were largely completed by the late 1970s. Bell Labs executives then began planning for a major installation of fiber along the northeast corridor. As for mobile phones, AT&T planned a rollout in several cities—pending a green light from the FCC. By 1980, the success of both technologies seemed assured.

  THE FUTURE OF THE BELL SYSTEM was far less clear. Between 1978 and 1980, the legal landscape for Ma Bell had grown increasingly grim. The government’s lawsuit asking for a breakup of AT&T had proceeded steadily forward. A new lawyer at the Justice Department’s antitrust division was now overseeing the suit—a Stanford academic, William Baxter, who promised to litigate it “to the eyeballs.”20 Also, there was a new judge hearing the case, a brilliant and mercilessly efficient former government lawyer named Harold Greene. Finally, the powerful and pugnacious John deButts had retired as chairman of AT&T. The new AT&T chairman was a mild-mannered electrical engineer who had begun his career at AT&T as a ditch digger and had worked his way up through the company via twenty-three different jobs over several decades.21 His name was Charles Brown. Everybody called him Charlie.

  One could see AT&T’s fundamental problem as arrogance. The phone company’s broad powers and influence, Judge Greene would later say, gave it “both the ability and the incentive to prevent competitors from gaining a substantial foothold.”22 This in turn had had a negative impact on consumers. During the trial, the judge would eventually hear a mountain of persuasive evidence put forward by government litigators; one after another, competitors would testify how Ma Bell had thwarted their efforts to enter the telecommunications business. At the same time, one
could view the trial as a great war over ideas. To William Baxter, the driving force at Justice, AT&T’s fundamental problem was that it was both vertically and horizontally integrated. Vertical integration meant that the company controlled its research, development, manufacturing, and deployment. One way to visualize the verticality was to see Ma Bell as a series of boxes, stacked one atop another, each representing essential units of the company. The bottom box, where ideas and innovations began, was Bell Laboratories. Above that was Western Electric, where those innovations were in turn mass-produced. Above that was the top box, AT&T, which deployed the new technologies in the long-distance and local markets.23

  Baxter did not object to the idea of vertical integration. It made sense that companies should fund research and bring the fruits of those investments to the market. It was the horizontal integration he considered unacceptable. Horizontal integration could be seen as a series of boxes, too, but these boxes stretched from side to side, or really from coast to coast, instead of from top to bottom. These boxes comprised AT&T’s long-distance group as well as the local Bell operating companies—New York Telephone, New England Telephone, Southern Bell, Northwestern Bell, and so forth. The government believed that AT&T’s control of almost all the local networks created a bottleneck preventing long-distance competition, such as MCI, from thriving. It also created a barrier to companies that wanted to build their own telecommunications equipment. As it was, Western Electric, a close corporate partner to phone companies around the country, had a near hammerlock on the telecom manufacturing business.

  The lawsuit was not, therefore, about Bell Labs. Executives at AT&T had nonetheless realized from the start that whatever fate befell the larger phone system would befall their laboratory, too. By 1980, Morry Tanenbaum, who invented the silicon transistor at Bell Labs years before, had become involved in the decisions over the company’s destiny. Tanenbaum had risen to become AT&T’s executive vice president, effectively making him one of Charlie Brown’s top deputies. “I was the only one of the senior officers who had spent time at Bell Labs, and Charlie was extremely concerned about Bell Labs,” he recalls. Brown worried that if he spun off Western Electric—the middle box on the vertical stack—then AT&T wouldn’t be able to take new discoveries from Bell Labs and move them up into the telecommunications network. Spinning off the local operating companies presented a similarly unappealing prospect: It would drastically reduce the funding available to Bell Labs, since those local companies represented a large portion of AT&T’s annual revenue.