Photonics Firms Continue to Invest in R&D

Brent D. Johnson

Optoelectronics companies are faced with the most difficult of challenges. They must continually design and build products for a world that doesn't yet exist. With any luck, they will emerge at the end of a five-year program with devices that are perfectly attuned to the demands of the market. However, there are countless imponderables that must be considered. Who could have predicted five years ago the demand for terahertz and thermal sensor technologies brought about by the war on terrorism, or the development of proteomics for that matter?

In the current economic environment, many companies accustomed to projecting their ideas years into the future are faced with the more pressing issue of immediate survival. The question for photonics companies is whether they will abandon the heroic experiments and intuitive leaps that have come to be a hallmark of the industry or favor a more conservative approach with incremental improvements to existing technologies.

Considering the liabilities and cost of taking an actual product to market, some companies view licenses as commodities. The question then becomes: Is your R&D program there to fight a rear-guard action, fending off competitors with patents and copyrights, or is it the spearhead of the company, aggressively making new inroads and opening up frontiers?

Larry Johnson, president of ILX Lightwave, said the only way to fight the gray market is to make things that no one else has. That means developing new products and funding a robust R&D program. Although his company has lost on the order of 75 percent of its income with the collapse of the telecom bubble, it continues to spend more than 15 percent of its revenue on R&D.

Johnson said you can push incremental performance issues only so far. In the early days, ILX made improvements to laser diode instrumentation and to noise and stability levels. Yet, at some point, you reach diminishing returns. He compared this to a car that can go from 0 to 60 in two seconds. Who needs it? At some point you need to start developing new capabilities, which is why approximately 10 percent of the company's R&D budget is spent on projects that are "speculative in nature" and that no one else is attempting.

Perhaps most striking are the figures offered by Corning Inc., whose recent achievements include the invention of Leaf fiber, 1737 glass used in flat panel displays.

In 2001, 80 percent of Corning's sales and revenue came from products that were less than four years old. This success has led the company to substantially expand its R&D efforts in the past five years. David L. Morse, division vice president and director of corporate research, science and technology, said, "Tough economic times are when [research, development and engineering] are most important and when they, in fact, can be most effective in making major advances."

So why fund an expensive R&D program when one could simply purchase the technology as it is needed? Certainly there are enough defunct start-ups to choose from, aren't there?

Randy Heyler of Newport Corp. acknowledged that it can be less expensive to purchase the rights to a proven technology than to develop an equivalent one internally. However, success can be circumstantial. "It may be more expensive in the long run to fund programs yourself, but if it's in an area of your core competency, you are generally better off doing it yourself." Newport's expenditure for R&D was 9.6 percent of sales for 2001.

Heyler said that Newport's percentage of R&D investment for fiber optic assembly and test systems this year is higher than average, even though the company has closed operations in some locations because of continuing investments across a lower revenue base. But absolute R&D spending in this sector is down substantially over last year.

The company now is focusing on submicron alignment and retention technologies for fiber or die bonding and on test and characterization systems for active devices.

Lynn Strickland, vice president of marketing and strategic development for Melles Griot, is pessimistic about the value of acquisitions as a means of technology expansion. "They can be long and painful things," he said. He believes that it is more important instead to invest in your own company. "If you buy a Skunk Works someplace, is that really doing anything for your core business or your employees?" he asked.

Melles Griot has made a significant increase in its R&D spending this year in both dollars and percentage. It is two years into a five-year development plan that is focused exclusively on core technologies. Previously, the company had 12 operating units, many of them pursuing unique R&D programs. These units have been funneled into three larger programs supporting optics and laser technology. This strategy of consolidating R&D efforts and partnering with allied companies appears to be one of the major industry trends.

John Roush, vice president of product management for PerkinElmer Optoelectronics, said that "Opto" comprises many businesses that are all spending on a local basis. Lately, they have adopted an integrated strategy that focuses on fewer projects and concentrates spending on "bigger opportunities." Amorphous silicon panels for digital x-ray applications is one area where the company sees a growth market. Digital projection lighting is another.

"When the economy is slow, you can't stop your product development," he said. PerkinElmer has modeled its R&D strategy on General Electric Co.'s multigenerational product planning, which forces managers to hammer out a road map that covers five years.

Roush is confident in the plan that his company has devised and believes that leveraging its core technologies, which include avalanche photodiodes, thermopile sensors and flash tubes, will allow it to serve both the consumer and corporate customers in military, biomedical, photographic and medical fields.

It seems clear that photonics can expect no shortage of innovations stemming from continued R&D investment. But will the customers be there in five years?

Only time will tell.