Closing the gap between design and production

DIFFICULTIES GETTING FROM THE DESIGN ROOM TO THE PRODUCTION FLOOR LIMIT SUPPLIES OF OPTICAL AMPLIFIERS, SWITCHES, AND OTHER ADVANCED COMPONENTS. THE FOUR MAJOR GATEWAYS THAT PRODUCTS PASS THROUGH TO MARKET ARE THE FOUR GENERATIONS OF R&D. JOHN GRADY

Research and development must result in highly profitable commercial products. Many companies that supply WDM networks will disappear if they can't quickly market reliable new products at competitive prices. But getting a product out of the research lab and into production can be a painful challenge.

How do optical component companies create strategies for manufacturing that will succeed? Whether it's wrestling with an uncooperative lithium niobate modulator or negotiating acceptable price points with the sales people, the process requires capable people who can juggle all the balls and keep a component moving smoothly from concept to reality.

The four steps to a finished product—concept, design, development, and prototype—remain the same, but the number of groups involved increases through R&D's four generations. First-generation research engineers are joined by manufacturing engineers in the second generation, sales and marketing people enter the mix in the third generation, and in today's fourth generation, customers are becoming involved as well.

"I can't get the components off the drawing board into production fast enough," says Philip Anthony, general manager and vice president for amplifier components at JDS Uniphase. It's a problem confronting the entire industry. And he's in a position to know—JDS rolled out more than 70 optical components so far this year.

One way to resolve these issues, which also flourished during the development of the semiconductor industry, is to turn to outsourcing, or contract manufacturing. "Manufacturing partner" is a business model slowly being introduced into the optical components industry (see this issue's "Perspectives," p. 29). Still, many component designers are tooling up their own cleanrooms and fabrication floors, investing in automated equipment, and ramping up for their own rapid manufacturing. As a result, optical component companies must search for a champion, the person who's able to take a device that five engineers can make work in a lab, turn the design around, and then produce it, in volume, in a hurry, with consistent and reliable quality.

"Who can solve all those problems?" asks Anthony. "Can you find an engineer who can do it all? A jack of all trades—that would be excellent—but such a person is very hard to find."

Solving these problems means understanding the processes involved, and identifying the break points. Many organizations and books deal with the time-tested process of research, development, and marketing (see "Further reading on the gateways to market," p. 44).

Basically, there are four gateways a new product must pass through on its way to market—concept, design, development, and prototype. Each has its own set of obstacles. The destination is volume production, but the research and development doesn't end there. Products are still refined, even as they make their way through the assembly line.

Research and development consultants discuss concepts that began in the 1950s as "first generation"—an age in which conceptual engineers ruled their own research domain. Technology managers determined concepts to pursue and timelines required. Connections between R&D labs and company plans were few.

Gradually, manufacturing concerns were integrated into the early research process—the "second generation" of product development. Marketing and sales professionals entered the picture in R&D's "third generation." Today's complicated development environment is the "fourth generation"—now customers are becoming integrated into the process.

By examining these R&D gateways, generations, and progressions, important new methods emerge that can make the difference in survival for optical component suppliers.

DIFFICULT PROCESS TO MASTER
Concept, of course, is simple. In some cases, it's been done on a cocktail napkin. But the probability of eventual success at this stage is low. In the good old days of "first-generation" marketing in the 1950s and 1960s, companies put their creative and independent researchers together in the labs and spent lavishly, hoping something useful would emerge. And if you were a Bell Labs or an IBM, then some fantastic—and very profitable—technologies did emerge.

In the next gateway—design—engineering teams must meet with product management and marketing teams to "cost out" their idea—a common area of failure. It may be a good idea, but who will buy it and at what price? This stage is when most concepts are dropped.

If the idea does move from concept to design, it also enters the "second generation" of research and development, the one in which activities become organized into projects, with progress measured against a set of established goals. Costs are carefully examined against potential benefits. Real customers must be targeted. Real connections to the company's overall strategy must be made.

If and when a concept survives this design phase and moves to the next gateway—development—the real problems begin. "This stage has the most delays, because all the problems must be faced head-on and solved here. You can't push too hard on a single aspect; there are just too many interdependencies," says Anthony.

Development must take place in the multidimensional world of third-generation R&D. Not only are the researchers and manufacturing managers trying to make a concept work, the company's marketing, sales, and executive branches also enter the fray. Here, projects must be integrated into the company's overall business plan.

One example of a great design that became stuck in development is the lithium niobate modulator. Engineers successfully made two parallel waveguides in this material that changed in refractive index with voltage. But when the design moved into production, the engineers found that the bias voltage started to drift—it could modulate at one voltage one time, and then require a different value the next time. This component remained in "development" for 20 years.

FOURTH-GENERATION STRATEGIES
The fourth gateway—prototype—can be the most difficult. And for some products, this is the final stage. Look, for example, at all the optical switches and crossconnects on display at trade shows—how many can really deliver what they promise?

Prototypes must be forged within today's "fourth generation" of research and development. The preceding three generations involved forces within the company; now the customer must be included to succeed.

In Fourth Generation Management, author Brian Joiner outlines an approach that transcends traditional methods and places the needs of the customer above all else. He recommends that long-term team-like relationships be formed with clients. And he expands the scope, saying managers must create an internal working environment where all employees can feel invested in the company's new products.

But potential customers often change their requirements by the time the prototype is built. A customer may say, "Sure I'd like a 2000 x 2000 optical switch." But when it comes to writing a check, well that's a different story. So the customer's needs must reach back all the way to the initial concept phase.

Innovative ways of addressing these issues are being created. William Miller and Langdon Morris, authors of Fourth Generation R&D, embrace principles of knowledge management as they blueprint new ways of achieving product success. They say the basic differences between explicit (written down) and tacit (experienced) knowledge can be exploited for R&D success.

A product's final gateway is production. Compared to the preceding development and prototype stages, it's relatively free of surprises. But the development process still doesn't end. "In production you get a product going but you don't stop trying to improve it," says Anthony. "You find a new problem and loop back to fix it."

The process of reaching this goal of profitable production is riddled with bottlenecks and obstacles. The person who can wear all the hats, keep all the balls in the air, and keep driving the product forward, no matter what, can make the difference.

"The main thing is, you have people involved at each and every stage," says Anthony. So that means it takes a person who can handle people to make the process work.

I asked the JDS vice president whether he is one of that rare breed who knows end-to-end product production, and is experienced the worlds of design engineers, manufacturing engineers, salespeople, marketers, executives, and, most important, customers. "No," he said, "but I am good at finding them and helping them do their jobs." "Oh," I said, "so then maybe you can give me a name and let me talk to him for my column?" He smiled and said, "No way."

Further reading on the gateways to market

The following resources may be helpful for those who are interested in reading more about R&D and manufacturing gateways to market.

Phil Roussel, Kamal Saad, Tamara Erickson, Third Generation R&D: Managing the Link to Corporate Strategy.

William L. Miller and Langdon Morris, Fourth Generation R&D: Managing Knowledge, Technology, and Innovation.

Brian L. Joiner, Fourth Generation Management: The New Business Consciousness.

R. K. Jain, Harry Charalambos Triandis, Management of Research and Development Organizations: Managing the Unmanageable, 2nd Edition.

F. Peter Boer, The Valuation of Technology: Business and Financial Issues in R&D.

Society of Manufacturing Engineers; www.sme.org.

Manufacturing Research Institute at Georgia Tech; www.marc.gatech.edu.

American Precision Optics Manufacturers Association; www.apomanet.org.

International Association for the Management of Technology; www.iamot.org.

Product Development and Management Association; www.pdma.org.

Industrial Research Institute, www.iriinc.org/web.

International Project Management Association; www.ipmacourse.dk/index.htm.