Champions of change
March 6, 2000
North Reading, MA--Early in 1996, Teradyne engineer Eric Truebenbach took on a task that he, like many engineers, detested. His task--writing a lengthy product specification for a new automated tester--is the kind that battles with the creative instincts of every engineer. It's a boring, detail-oriented job that can take weeks. Excruciating weeks.
But as Truebenbach sat down to write this particular spec, he made a surprising discovery: This one was easy. For five days the words tumbled forth, until Truebenbach had a 40-page treatise. And those 40 pages were more coherent, more concise, than any spec he'd written in his 13 years with the company. "Ordinarily, the worst part of any project is the specification," Truebenbach recalls. "It's like giving birth. There's a lot of huffing and puffing. But that didn't happen here. We had all the data. There was no stress, no sweaty palms."
Truebenbach says that the product spec was easy for a good reason: He could already see the tester, in all its electronic detail, in his mind's eye. "If we would have had ten engineers right then, we could have built the whole system," he says.
How is such assurance possible so early in a project? The answer is as unlikely as it is simple: His vision was inspired by a quality program--one that took Truebenbach through a journey that involved hundreds of meetings and countless hours of give-and-take with marketing and manufacturing. "Before I sat down to write the spec, we went through two-and-a-half months of negotiating," Truebenbach says. "And when we finished, we knew what we were doing and why. We knew which items were important and which we could trade off."
Teradyne executives then combined that inspired vision with an intense development program in which Truebenbach and his engineering colleagues evaluated, re-evaluated, and re-re-evaluated the design of the new tester. Result: the M910, a digital test instrument that so impressed one customer that the company's executives asked Teradyne to help with their quality program.
Such requests are now growing increasingly commonplace at Teradyne. In the world of test equipment, the company's reputation for quality is almost legendary. It's been ranked among the ten best test equipment manufacturers for 11 straight years by VLSI Research Inc., a San Jose-based firm that surveys the semiconductor industry. It has also been named as one of America's Elite Factories by Fortune magazine and as the 1996 Company of the Year in Massachusetts by The Boston Globe.
And on the strength of the M9-Series, as well as the company's showing in the semiconductor industry, Teradyne is this year's recipient of the Design News Schneeberger Quality Award.
Corporate turnaround. Teradyne's fortunes, however, weren't always so bright. During the early 1980s, there were times when its future looked positively dim. "We kept hearing from customers that Japanese quality was better than ours," recalls George Chamillard, now president and chief executive officer of Teradyne. "Their machines were getting a thousand hours mean-time-between-failure, while we were getting about half of that."
For Teradyne's executives, such revelations came as a shock. The company's top people had, after all, prided themselves on their corporate culture. The firm was run by engineers--people who knew how to build automated testers from the ground up. And its corporate culture was flat. The top officers--all the way up to and including the CEO--sat in cubicles. There were no executive washrooms. No executive parking spaces. No mahogany row. No titles on their business cards.
But some issues were out of their control. Much of the semiconductor manufacturing market was moving to Japan. So Teradyne struggled to maintain its market share. The company lost money in four out of five straight years. "We were really under siege," Chamillard remembers. "The entire industry was in crisis."
Like many American counterparts, Teradyne executives decided to study quality. High-level managers were told to read six books about the subject, starting with W. Edwards Demings' classic Out of Crisis. All the managers read the books in the same order, and then each attended at least two quality-related seminars. At the same time, Teradyne also signed on as one of seven charter members of a Boston-based group called the Center for Quality Management, which included Bose, Polaroid, Digital Equipment Corp., Bolt Beranek and Newman, Analog Devices, and GE Aircraft Engines Division.
Slowly, the firm's executives began to learn about quality manufacturing practices. By 1990, they had started an internal TQM (Total Quality Management) office. Still, that wasn't enough. The firm's founder and chairman, Alex d'Arbeloff, wanted to extend the quality efforts to the engineering practice. A former engineer, d'Arbeloff felt that a manufacturing-only program would treat the symptom, not the cause. "Alex was the biggest crusader in terms of pushing the design process," Chamillard explains. "He felt that most of the issues in manufacturing start with design."
Still, the company's executives had no idea how to apply TQM to engineering until a couple of them stumbled on an executive program at Harvard Business School. The program, based on a book called "Revolutionizing Product Development," held that there were two key paths to engineering quality. The first was choosing the right projects. The second, executing them properly.
Teradyne executives were so impressed with the course that they hired its instructor, Harvard professor Gary Pisano, as a consultant. Pisano launched his consulting effort by interviewing 60 people from the company. His conclusion: that Teradyne's engineering ranks were approximately 200-300% over-committed. "It's common among manufacturing firms," Pisano explains. "The engineers had too many projects. And the projects were taking too long. So the company moved people from project to project, and it ended up affecting personal productivity."
Many firms, Pisano says, try to solve that problem by adding more engineers. That, however, seldom works. "That's like buying a bigger pair of pants without going on a diet," he says. "Pretty soon, you're going to need a bigger pair of pants again."
So instead of telling them to buy larger trousers, Pisano placed them on the engineering version of a diet. He suggested they build new products off a common platform, thus leveraging as many of their resources as possible. He taught them to make sure all new products were consistent with each division's strategy. And he implored them to develop a structured process for putting new projects on the agenda. That way, the project list would match the company's available manpower.
At the same time, Teradyne formed an Engineering Process Improvement Team that met eight times a year in locations around the country. Known as EPIT, it included engineering heads of each of Terdayne's seven divisions, its service organization, and foundry. Pisano also attended EPIT meetings. Together, the group hammered out official documents for Path One (choosing the right projects) and Path Two (execution). Each document consisted of approximately 60 loose-leaf pages in a binder. "The documents aren't cook books," notes Teradyne vice president Ed Rogas. "But they establish a framework. And they make our people understand that they have to set aggressive goals and execute with precision."
Executing the strategy. None of that would count for anything, however, if engineers weren't able to translate the theory into better products.
And that's what Eric Truebenbach set out to do. Truebenbach, who joined Teradyne as a hardware designer after graduating from Tufts University with a BSEE in 1983, had previously been the architect for significant portions of the company's L200, L300, and M895 test instruments. Colleagues say he was an ideal candidate to carry out the new philosophy of Revolutionizing Product Development. "When it comes right down to it, you need more than theory," notes John Wood, engineering manager of the Assembly Test Division at Teradyne. "You need someone to be a champion of change. And Eric was our champion of change."
At the beginning, Truebenbach was alone on the M910 project. And his task was formidable. With the M910, Teradyne intended to change the way they built automated test equipment. Up to that time, the company's test equipment had used proprietary hardware and software. During 1996, however, Teradyne executives saw a need for a system that would meet industry VXI standards, which call for certain standardized sizes, shapes, and connection configurations of various parts. "We were finding that some people didn't want those big, proprietary integrated testers anymore," Truebenbach recalls. "They wanted more flexibility so they could add and subtract whatever they wanted."
What's more, the company knew that product reliability would be a key. Predecessors of the M910 had been used to test avionics boards on aircraft carriers. And the first customer of the new product would be the U.S. Marine Corps, which planned to use it onboard field vehicles. "Reliability is critical in those situations," notes Maggie Cadogan, Teradyne's companywide TQM manager. "If a part breaks in the middle of the Indian Ocean, you're not going to get Fed Ex to deliver a new one overnight."
Achieving higher reliability, however, wasn't going to be easy. The system was not intended for use in lab-type settings. So instead of facing temperatures ranging from 25 to 35C, as a bench-top instrument would, it would be subjected to a range of -10 to 50C. To make matters more difficult, cost goals called for a 50% reduction from the previous generation. And the targeted reliability--10,000 hours mean time between failure per VXI card--was about five times higher than previous models.
Power and size. To meet all those requirements, Truebenbach concentrated on the issues of power, size, and cooling.
To start, he reduced the voltage range of the circuit boards. Instead of a range of -5 to +15V, as previous systems had used, he decided that every component would have to operate between -2 to +5V. He also incorporated a power-down mode and a more efficient power-conversion technology. Then he revised some of the ASICs (Application Specific Integrated Circuits) to minimize their voltage draw. By doing all that, Truebenbach hoped to reduce power consumption and, therefore, the size of the circuit boards.
By themselves, however, those changes weren't enough. So to further reduce power consumption, Truebenbach made an enormous spreadsheet showing each of 1,300 components in the system, along with their power consumption. Then he consulted with marketing and went back through each of the components, eliminating anything he could. "Every time marketing said we needed to reduce power, I went back over my spreadsheet and tweaked the design wherever I could," Truebenbach recalls.
To further reduce size and weight, he worked with fellow engineer John Thorp, one of 18 engineers who ultimately ended up working on the project. Truebenbach and Thorp examined component sizes and material choices. They took each component down to a scale in the building's mail room, weighed it, graphed the results, then made changes.
As the M910's size shrunk, however, it presented Teradyne's team with another design dilemma: cooling. Somehow, they needed to find a way to dissipate the heat caused by the tighter packaging densities. To do that, Truebenbach teamed with Teradyne engineers Mike Lozan and Gene Veilleux. Together, they built a mechanical mock-up of the system, blew air across it, and calculated the pressure losses. They analyzed the system in a thermal chamber, designed new heat sinks, and rearranged parts on the board so that the hotter ones were subjected to more air flow. They even used components of varying heights, thus creating a little "city skyline" that more efficiently channeled air cross the circuit boards.
In the meantime, software engineers Alycia McGoldrick and Teresa Lopes developed a Windows NT-based software driver to operate the instrument's hardware. The NT-based drivers were a concession to the changing nature of the automated test market, which had begun to call for open non-proprietary systems.
While all this was going on, Truebenbach continued to attend weekly status meetings. There, members of the quality team grilled him about the design choices. The questions were seemingly endless: What process did you use to come to this conclusion? Where did you get your data? What are your alternatives if it doesn't work? "I dreaded the weekly status meetings," Truebenbach recalls. "Yet I knew they were helpful. There is no tougher boss than one who truly understands quality management tools."
Walking the talk. The result of their efforts, the M910, met the company's goals. Engineers were so successful in reducing the instrument's size that two of its channel cards replace 10 previous cards measuring 15 X 20 inches. Cost of the entire M910 is around $100,000, compared to a minimum of half a million dollars on earlier systems. And engineers met their goal of 10,000 hours MTBF--thus quintupling the figure of earlier systems.
The initial M910, introduced in September, 1997, was incorporated in the Third Echelon Test System (TETS) designed by ManTech Systems Engineering Corp. (Chantilly, VA). TETS tests the performance of signals from electronic systems onboard tanks, radios, missile launchers, and other military field equipment. Since its introduction, Teradyne has sold M9-Series instruments to all branches of the military and to commercial customers, as well.
They have also leveraged their efforts on the M910 in the design of its two successors, the M920 and M960. By doing so, they cut costs, strained their resources less, and ensured that they were staying within the strategic vision of Teradyne's Assembly Test Division.
Most important, though, is the cultural change that the project brought to Teradyne. Engineers at Teradyne's facilities speak their own language of quality (see sidebar). And they now teach their quality system to customers and vendors who need help with product reliability. "I've worked with a fair number of companies," notes Pisano of Harvard. "And Teradyne is unique in the sense that they really institutionalized their quality effort. They built it into their language and their thinking. They don't just say the words. They walk the talk."
Quality speak
Spend some time with Teradyne's engineering staff, and you're bound to hear a few baffling phrases. Why? Because the company's quality effort has produced a language of its own. Following are examples.
Quality speak | What it means |
---|---|
"Jump up." | Let's look at the big picture. |
"Jump down." | Take a narrower view of the topic. |
"Jumping to step 4" | You're jumping to conclusions. |
"Spin a QIT." | Check out a possibility. |
"Shoot the fat rabbit." | Look at the big bar on a bar chart. |
"Funnel it." | Select the best idea. |
Dealing with overload
When Harvard Business School professor Gary Pisano analyzed Teradyne's product development cycle, he found that the engineering work force was overloaded. Such situations are very common among manufacturing firms, he says. Here's what Pisano recommends for companies that want to get out from under the overload.
1. Leverage your resources. Rather than trying to develop three independent products, try to use one as a technology platform to build the other two. That way, you use your engineering resources more wisely.
2. Link corporate strategy to individual products. If a new product doesn't fall within a company's strategy, then why build it? Companies that make this mistake usually have not bothered to articulate their strategy, Pisano says. And if a firm hasn't articulated a strategy, then its executives often won't know if a new product idea is appropriate.
3. Create a formal process for putting new projects on the agenda. This forces engineers to think about the company's goals and strategies when they come up with a new idea.
Revolutionizing Product Development --the book
When corporations adopt quality programs, they typically focus on manufacturing. When Teradyne launched its quality program, however, it went one better. Instead of simply focusing on manufacturing-based reliability issues, the company's executives decided to apply quality tools to the engineering process.
Their initial source of knowledge on the subject was a book, Revolutionizing Product Development, written by Harvard professors Steven C. Wheelwright and Kim B. Clark. The book, published by The Free Press, is based on the idea that if companies bring products to market faster and better, they'll be more successful. It cites scores of successful examples, ranging from Honda in the automobile industry to Applied Materials in semiconductors to Sony in audio equipment.
Teradyne engineers used the book to create a framework for product development. Their framework is based on the book's two paths: choosing the right projects and then executing them properly.
Initially, Teradyne applied the book's principles to solve the problem of an overloaded engineering staff. Using a concept called "the development funnel," they were able to identify the most promising concepts for development, and avoid those that didn't match their strategic objectives. Result: They winnowed down the workload, and got products out the door faster.
But while the book's concepts may be easy to understand, they take time to learn. "It's not a cookbook," notes Gary Pisano, a professor at the Harvard Business School. "If you're going to use it successfully, there are hard decisions to be made. But if you're willing to make some behavioral changes, you can get more bang for your buck in the product development process."
About the Author
You May Also Like