Next month, Brad Parkinson receives the IEEE Medal of Honor for his work bringing GPS navigation from concept to reality. Parkinson told me in January that he’d always wanted a cheap, consumer version of the technology—though he never imagined just how cheap and ubiquitous it would become.
Writing a profile of Parkinson sent me down a trail of my own—into my memory banks, to recall the others who made GPS something we can take for granted today. In addition to Parkinson, there’s Ivan Getting, whose vision set the stage for GPS today—the full text of my 1991 article on Getting is here. And there’s Charles Trimble, who shared Parkinson’s vision of consumer applications for GPS. His company, Trimble Inc., introduced its first GPS product in 1984, a system designed for off-shore oil drilling platforms. In 1994, the company brought out its first GPS receiver that fit on a PC card. The full text of that profile is below.
Charles R. Trimble
This former HP engineer started his own company to exploit the newest free resource—the Global Positioning Satellite system
In 1982 the U.S. Global Positioning Satellite program—a U.S. government plan to launch a constellation of satellites that would allow users worldwide to pinpoint their locations—was by no means a sure bet. Receiver technology was only in the research and development phase, and government budget slashing was already beginning to eat away at the project’s funding. That year Palo Alto, Calif.-based Hewlett-Packard Co., which had had four engineers working on the technology for some four years, took its cards off the table.
Whereupon Charles R. Trimble decided to get into the game. A former HP engineer then leading a struggling start-up company, he put all his chips on the Global Positioning Satellite (GPS) program, and in spite of near disaster—the Challenger explosion put the GPS program on ice for three years—the gamble proved a winner.
And those winnings ballooned into a bonanza during the Persian Gulf War, when the U.S. military turned to Trimble Navigation Ltd. For some 10,000 portable GPS receivers. That kicked the expanding company into overdrive, and it is now growing at over 100 percent annually. From last year’s revenues of US $63 million, analysts project 1991 revenues will be $145 million. The company currently has 750 employees, including some 200 engineers, an increase of 250 workers over 1990.
A quick tempo appeals to Trimble. Though a native Californian, he moves like a New Yorker late for a train, rushing from building to building in the Sunnyvale, Calif., industrial park that houses his company. In his career, too, he has always been a mover and shaker: in 14 years at HP, he was involved in four new ventures, leaving only when that company’s intrapreneurism slowed down.
Trimble had intended to work as an engineer for only three months. In 1964, he had gained a master’s in electrical engineering from the California Institute of Technology in Pasadena—which he picked in junior high school after hearing it was the hardest school to get into—and was planning to enter Harvard Business School in Cambridge, Mass., in the fall. Meanwhile, he took a summer job at HP.
“I viewed that as my only opportunity to understand the engineering world,” Trimble told IEEE Spectrum. “So I put my heart and soul into it, working as many as 90 or 100 hours a week.”
Before that summer, he had been strong on scientific theory, but had never even picked up a transistor. By Labor Day he was hooked on engineering, and called Harvard to withdraw.
Trimble blames that move on the man who hired him, Al Bagley, then head of HP’s Frequency and Time Division, and his determination to keep Trimble out of Harvard’s grasp.
Nor does Bagley deny the charge. When recruiting at Cal Tech that spring, he had heard about “this kid Trimble” from everyone—from professors to the clerks in the bookstore. The student was impressive in an interview, Bagley recalled, but did not want full-time employment. Bagley’s bait was a challenge: could Trimble build a “computer of average transient”—a digital device that could analyze signals and spot repetitive signals amid noise? If he succeeded, Bagley promised that HP would market the creation as a special-purpose instrument.
The new graduate thought the project would take two years, but it took four, and his plans for business school faded. The HP 5480 Signal Averager bowed in the fall of 1968 and became primarily a biomedical research tool.
If Bagley had had any doubts about his new hire’s intrepidity, they dissipated one stormy day during a sailing race on San Francisco Bay. Safe on shore, Bagley saw masts breaking and on one small sailboat, a wildly swaying mast with Trimble perched atop.
“That is the sign of a gutsy guy,” Bagley said, particularly in light of Trimble’s congenital eye defect. Scar tissue on the retina deprives him of all central vision He therefore cannot drive and must read through binoculars or by holding documents close to the periphery of his eyes.
INTRAPRENEURISM. When this “gutsy guy’s” computer of average transient was completed and on the market in 1968, he asked for the chance to figure out a major new business area for HPs Frequency and Time Division. Bagley agreed, and Trimble targeted real-time dynamic testing of large-scale integrated (LSI) circuits. Assembling a team of 20 people, one of the largest groups in the laboratory, he began building a computer-controlled dynamic IC tester. Though static testers of LSI chips existed, no dynamic testers were then available for that level of circuit integration.
In 1971 the prototype was completed, but money looked tight for HP for 1972, and Trimble’s project was canceled.
Then, undaunted, he started on another new project—focusing on single-shot time interval measurement and connectivity between electronic instruments and computers and calculators. One of his project managers became the instrumentation point man in the interfacing effort that led to the HP Interface Bus (HP-IB) and the IEEE 488 standard.
Trimble’s last job at HP was as head of development for the Santa Clara-based bipolar LSI laboratory, which basically did contract engineering for other company divisions. Such a structure allowed for little discretionary work, and Trimble was interested in using the IC fabrication line to develop a new series of instruments capable of measuring time intervals as short as 50 picoseconds. So he persuaded corporate management to restructure the laboratory: in return for a fixed budget instead of total dependence on contract research, he would reduce staff and eliminate efforts in publishing and giving academic papers, which were taking up one-third of researchers’ time.
Shaping the HP way to suit his goals was standard operating practice for Trimble, according to Jim Sorden, an HP colleague who is now vice president of product engineering at Trimble Navigation. “It was well known in HP that if you wanted to beat the system, talk to Charlie, and he would figure out the way,” Sorden said.
When that project ended in 1978, Trimble looked around HP and realized that the environment had changed. The price for entrepreneurism by that time was working on projects that were 10 years from production. “HP had gone from a technology-driven to a marketing-driven to a resource-allocation company in my 14 years there, Trimble told Spectrum. It was time to go.
Colleague Sorden was not surprised. “Though Trimble did well at HP,” he said, “when you get the reputation of a rebel, you are clearly not going to move up the ladder in a company that is becoming more and more conservative.”
BOOTSTRAPPING. Trimble then turned his attention to an HP development project in Loran C navigation equipment that had been canceled. Loran C uses time differences between low-frequency radio signals from land-based transmitters to guide navigators primarily on the U.S. coastline and the Great Lakes. Trimble approached the project’s division manager and, after several months of negotiation, purchased the Loran technology for $50,000, funds he obtained through refinancing an apartment building he had purchased earlier. He left HP with two engineers from the project, Thomas Coates and Daniel Babbage, and an administrator, Kit Mura-Smith, and set up shop as Trimble Navigation Inc. in November 1978.
Though he had always been a hard worker, Trimble was in for a few surprises when he took his entrepreneurial talents outside of the sheltered corporate environment. “I thought I knew what hard work and commitment was,” he said. “But I was totally unprepared for the set of emotions and stresses involved in trying to bootstrap a business from zero.”
On his own, survival was the No. 1 concern. Trimble said he was lucky to get through the early days: “I’m at a disadvantage in the $0-$2 million business level—I’m a strategist, not a street fighter.”
By 1982 Trimble Navigation was selling about $1 million of Loran equipment annually. But growth in the Loran market was flat, so he began looking for a new business area to pursue. He found it again in the reject pile at HP.
The company had canceled a development program for navigation products that used the nascent Global Positioning System. AT the time, plans called for GPS to be completed by 1987.
Trimble, enthralled with the idea of GPS as a free information utility ripe for commercial exploitation, bought the rights to an HP GPS breadboard in the summer of 1982. With the help of HP’s original design team, doing after-hours consulting, he drafted a fundamental GPS block diagram, which he considers his best piece of technical work to date.
But Trimble lacked the time to work through several generations. He needed a product he could sell within 11 months, at which point, he calculated, the company would run out of money.
The product area he chose was an obscure one—time calibration of cesium clocks, using atomic clock signals from PS satellites to calibrate the clocks at naval observatories—at best, a $1-million-a-year market.
Trimble knew that once the GPS constellation was complete, his $1.8 million company would have to have grown up, with revenues of at least $50-$100 million annually, because large competitors would then jump into the business. “We needed the resources to play in the end game,” said chess player Trimble. The company required growth of 60 percent a year, and he consistently met and exceeded that.
In 1985 Trimble Navigation introduced GPS positioning for offshore oil surveying and developed a GPS navigation sensor for aviation (of use to pilots when the GPS constellation was complete).
Then in January 1986, with only seven GPS satellites in orbit, the Challenger blew up. GPS launches were put on hold until a new rocket booster could be developed for GPS satellites—a delay of three years.
“With the satellites up there, we could survey to 25 meters, which was good enough for offshore use,” Trimble said. “If one satellite died, we could still survive. If two died, we would have been in deep, deep trouble.” Meanwhile, the collapse of oil prices hurt the offshore survey marketplace.
So the company began looking for customers that were unconcerned about the number of satellites in the sky because they were confident of GPS’s efficacy in the future. The prime candidate was the U.S. military, but the military had let an exclusive end-user receiver contract to Rockwell International Corp.-Collins Avionics and Communications Division through 1992, covering all the GPS applications it could think of.
The task for Trimble was to figure out applications Rockwell and the military had overlooked by assuming receiver equipment must cost a lot. He came up with a GPS brain for remotely piloted vehicles and personal position finders for the infantry, and soon created a small rugged finder that could be carried in the hand, in a pouch at the waist, or around the neck. The military took the bait, ordering 1000 Trimpacks at $4000 each, deliverable in May of 1990.
GULF BONANZA. When Saddam Hussein invaded Kuwait, the U.S. military began escalating its orders. Another 1000 unites were purchased, then nearly 9000 more. Trimble went from a company shipping $5 million in products a month to one shipping $19 million a month just six months later—all the while hanging by its nails as components grew ever harder to obtain.
The only shadow on this success story is Trimble’s dislike of being thought of as a military contractor His is primarily a commercial and consumer products business, and he wants to keep any military manufacturing down to less than 20 percent.
Trimble cut his management teeth at HP and has, he said, managed the HP way ever since He defines this managerial style a one of trading autonomy for commitment.
But autonomy does not mean hands off. On a summer’s morning as observed by Spectrum, Trimble sprinted from meeting to meeting, carrying no documents and taking no notes, but seemingly up-to-speed in every area—from review of current chip designs in development, to government lobbying concerning export controls, to personnel requisitions, to new product proposals, to plans to push defects closer to zero. Mostly he listened, occasionally he asked questions, and often he stumped his engineers.
Modeling complex problems simply so they can be solved by himself or others—be the problems scientific, engineering, or economic—is one of Trimble’s strongest talents. In fact, he said, detail bores him. Ralph Eschenbach, vice president of avionics and sensors at Trimble Navigation, thinks this aversion to detail relates to Trimble’s eyesight, which identifies overall shapes better than fine detail. “He tends to force things into simplified models and doesn’t get trapped with details,” Eschenback said.
Trimble himself attributes this skill to his family’s practice of doing mental math games on long car drives, and to Cal Tech professor R. David Middlebrook, who insisted that students solve a problem on a single page.
These days, he said, his job is “to gather people around me who are far more talented than I am. It is their turn to generate the inventions.”
TO PROBE FURTHER. Trimble Navigation Ltd. Has published an explanatory guide to Global Positioning System (GPS) technology, called “GPS, a guide to the next utility,” (1989).
“Navstar: the all-purpose satellite” [IEEE Spectrum, May 1981, pp. 35-40] discusses the technology behind GPS and the original intentions for such a system.
Originally published in IEEE Spectrum, February 1992, pp. 46-48.