Remarks

Dr. Arden L. Bement, Jr. Director National Science Foundation Biography "Government's Role in Industrial R&D" AAAS Annual Meeting Symposium on Industrial Research Trends Washington, DC February 18, 2005

Good morning. I'm glad to be with you to share in these discussions. I want to especially thank Neil Paton for inviting me to participate in this symposium.

When considering my topic, government's role in industrial R&D, it seems reasonable to ask, "Does government have a role in industrial R&D today?" After all, Federal funding of R&D has steadily declined since the mid 1980s. Today, the government supports about 28 percent of total national R&D¹. Conversely, industry's investment in R&D has increased to roughly two-thirds of the country's R&D expenditure. NSF's latest Science and Engineering Indicators shows that in 2002, industry performed 70 percent of the research and development in the United States. Looking strictly at these numbers, it might seem that government's impact on industrial R&D is waning.

Now, let me share another statistic. While public funding of R&D is 28 percent, the government provides almost 60 percent of the funding for fundamental research²--the basic, often ground-breaking, work that paves the way for industrial R&D. Examination of patent citations shows that peer-reviewed science journals are playing an increasingly important role as "prior art" in patent applications.

In short, publicly-funded research is contributing more and more toward industrial innovation³, giving government an invaluable role in industrial R&D. In fact, I see three roles: government as enabler, as crusader and finally, as cultivator.

The Federal government enables modern industrial research and development by creating technologies that sustain a stable R&D enterprise. Greg Tassey, a senior economist and colleague at NIST, calls these "infrastructure technologies" or "infratechnologies." Like any utility, infratechnologies are usually overlooked--the unsung heroes of R&D. That is the bane of a reliable infrastructure--no one notices until it breaks.

Take the electrical grid, for example. When you turned on the light this morning, did you marvel at the efficiency of our power distribution network? Neither did I. Most people don't think of the electric company until they get their monthly bill. Be sure to send a "thank you" card with your next payment.

R&D infratechnologies are likewise taken for granted, yet they are vital to industrial research and development. From test methods, to process models to manufacturing and measurement standards, infratechnologies increase productivity, decrease costs, and open doors to new opportunities. I'll use metrology, the scientific study of measurement, as an example.

Metrology is indispensable in today's R&D environment. Measurement standards are essential for commerce. You can't sell a pound of anything until both you and your customer agree on what constitutes a "pound."

Metrology's importance has expanded beyond the weight of produce, however.

For example, nanotechnology promises to be the next frontier for industrial innovation. Researchers are regularly creating new nanostructures and discovering unique applications for nanoscale materials. The next challenge is to turn these materials into products that meet real needs. We need new metrology tools that enable practical fabrication at the nanoscale.

One hundred years ago, automakers were frustrated by mis-measured parts on an assembly line. Today, the challenge is to fit nano-scale gears onto nano-scale axles while attaining efficient production. The parts of a nano-machine can't interact without accurate, reproducible measurements.

One researcher puts it bluntly, saying that the nanotechnology revolution is "dead in the water without a metrology infrastructure."

Semiconductor manufacturing also relies on accurate, standardized measurement. A 1998 NIST study estimated that the U.S. semiconductor industry spent about $7.6 billion last year on measurement. That expense could stymie growth in the industry. The Semiconductor Industry Association recently declared, "One of the biggest barriers to continued miniaturization... is the inability to measure... critical dimensions at the desired... size." The Association concludes, "...additional research in metrology... is critical to future chip development." Accurate measurement, and the commensurate expense, is a necessary evil that government is working to resolve.

Government's second role, the crusader, is about confronting necessary evils while challenging the status quo. We must constantly advance science by funding risky, but potentially rewarding, research. Government also has an obligation to tackle problems that are important to society, but may have little commercial value. The tsunami disaster is a sobering reminder that we are at the mercy of nature.

Currently, geologists rely on ultra-long period seismometers to predict an earthquake's ability to create tsunamis. These sensors are an essential component of the Global Seismographic Network; however, the flagship ultra-long period seismometer is obsolete.

The market for these super-precise instruments is small, and the cost to develop a new model is high--estimated at $10 to $20 million dollars. The private sector cannot afford to pursue this project. Enter the National Science Foundation; we must take the lead in funding development of this new sensor. While we are building a better seismometer, we will also generate new knowledge with new industrial and societal applications.

Industry ultimately exists to make a profit--it is one of the tenets of our economy. Companies are not inclined to embrace too much risk; they have to realize a certain return on the investment. The Federal government doesn't have that constraint. Federal science funding realizes profit in a different way--acquired knowledge leading to innovation.

The riskiest research endeavor can yield breakthroughs in unlikely, seemingly unrelated, fields. An obvious example is the space program. Many thought that President Kennedy's plan to visit the moon was a waste of taxpayer's money. Yet today, we still reap the benefits. Our quest for the moon modernized aerospace engineering, materials manufacturing, medical sensing and countless other industries.

The National Science Foundation occasionally draws fire for high-dollar spending on risky projects. The Laser Interferometer Gravitational Wave Observatory, or LIGO, is a good example. LIGO is designed to detect gravity waves--ripples in the fabric of space-time.

Einstein predicted their existence in 1916, but so far, no one has proven that gravity waves exist. Constructing LIGO is a risk--it certainly wasn't cheap--but it has a huge payoff. After gravity waves are identified, we can test many aspects of the theory of general relativity. For example, LIGO will help us verify the existence of black holes. It will give us new insight into the nature of neutron stars--the collapsed cores of supernovae.

So, how does LIGO contribute to industrial R&D? As the road to the moon was paved with industrial innovation, likewise, building LIGO has driven improvements in laser optics, digital control systems, seismic dampening and other technologies. These new technologies, and the knowledge created in their development, are destined for future industrial products and processes.

The creation of new knowledge leads me to the third role of government in industrial R&D--that of cultivator. The government is uniquely suited to "grow" new researchers through support of academic and small business R&D. NSF has several exemplary programs, but I want to focus on two: the Small Business Innovation Research program and NSF's Engineering Research Centers.

The Small Business Innovation Research program, better known as SBIR, is a Government-wide initiative spread over ten agencies, including the Departments of Agriculture and Defense, along with the Environmental Protection Agency, NASA and NSF. SBIR awards focus on commercializing research results--turning out an innovative, viable product that people will want to buy.

SBIR success stories are plentiful. For example, Vanu, a small company based in Cambridge, Massachusetts, developed software that routes cellular telephone signals as effectively as specialized equipment, yet the application runs on a desktop personal computer.

Since June 2003, Mid-Tex Cellular has used the product to provide mobile service to several rural Texas towns. Large carriers are also interested; the Software Radio and its modest hardware requirements will reduce equipment costs.

Vanu's product may enable carriers to replace a room full of commercial routing gear with industry-standard PC hardware.

Another SBIR creation, FAST-ACT, is a nano-engineered product that neutralizes a variety of chemical warfare agents and hazardous chemicals. The material is fast-acting, able to remove over 99 percent of a compound in less than 90 seconds. These capabilities make FAST-ACT especially useful to hazmat teams and first responders, giving them a single technology that quickly counteracts numerous toxic substances.

To promote these and other successes, NSF sponsors regular "match making" events between SBIR grantees and venture capital and industrial firms. SBIR program manager Joe Hennessey calls this a "small business dating service" that introduces small business owners to strategic corporate partners. SBIR also has its share of academic connections, with about 75 percent of SBIR companies originating from a university incubator.

NSF's Engineering Research Centers, or ERCs, further emphasize the synergy between government, industry and academe. Industry is actively involved, whether through joint research, student mentoring or proof-of-concept testing. All of the ERCs have an Industry Advisory Board that reviews center activities and recommends new projects. Unlike the SBIR program, technology developed at an ERC doesn't normally pass directly from the center into commercial applications. The researchers develop an idea to its proof-of-concept phase, and transfer the technology to industry for refinement and commercialization. Perhaps the most valuable aspect of our centers is that they serve as a talent pool for industry.

One of the first ERCs, MIT's Biotechnology Process Engineering Center (BPEC), boasts alumni in every major pharmaceutical company as well as in several leading biotechnology firms. Graduates working at Genentech, a biotech company in San Francisco, are helping to bring new cancer treatments to market. These new drugs kill cancer cells while reducing the side-effects of traditional chemo-therapy. Center alumni helped develop new processes to manufacture the drugs, and even played a role in designing facilities to produce the new treatments.

In addition to bringing skilled talent to the industry, the Engineering Research Centers serve as incubators for start-up companies. Start-ups provide an important boost to the industrial research base because they can take early R&D risks that larger firms often avoid. To date, 104 new companies, employing nearly 1,200 people, have "spun-off" from NSF Engineering Research Centers. These companies are often founded by ERC faculty members, recent graduates, or students.

DigitalPersona, for example, was launched in 1996 by two graduates of the Center for Neuromorphic Systems Engineering at CalTech. Specializing in fingerprint identification, DigitalPersona's first product, "U. are U.," won the 1997 Best of Comdex award. Today, the company sells a complete line of fingerprint identification products, including "Password Manager." "Password Manager" allows computer users to associate passwords for e-mail, Web sites, and bank accounts with their fingerprint. Using "Password Manager," users can log into computers, Web sites or e-mail with the touch of a finger. Microsoft selected "Password Manager" to use with its line of fingerprint readers.

Another ERC, the Integrated Media Systems Center, or IMSC, at the University of Southern California, has produced nine start-up companies so far. The IMSC, NSF's elite center for multimedia and Internet research, also showcases the international partnerships we create. IMSC's industry members include Phillips Research, based in the Netherlands, ST Microelectronics, from Italy, and LG Electronics, a Korean company. Japanese telecom giant NTT DoCoMo joined the center this year. Domestic partners include Eastman Kodak, Hewlett-Packard and Lockheed Martin.

All of these examples illustrate that Engineering Research Centers are not only "greenhouses" for growing top-tier researchers; the centers are also stimulators for economic growth, both domestically and abroad. The international connections we forge are becoming more critical. Just as technology drives the growth of the U.S. economy, it can spur growth in less-developed nations, too. Ultimately, this will open new markets for U.S. companies.

Economic studies show that technology innovation comprises three-quarters of the productivity growth, and one-half of the gross domestic product growth in the industrialized nations. Since the mid-1990s, the preponderance of U.S. productivity growth is due to education and technology investments. Put succinctly, investments in education and science and engineering drive the economy. Likewise, the Federal government, in partnership with industry, must continue to drive fundamental research in science and engineering.

Government support of industrial R&D is a major force in pushing both science and the economy. Government enables industrial research and development by creating the infratechnologies that underpin R&D. We also work as a crusader--constantly pushing toward new frontiers, with an eye out for both industry and society. Finally, government is a productive cultivator--growing and nurturing future generations of industrial researchers, and seeding the economy with new companies pushing innovative ideas. I hope all of us leave today's discussions with a better appreciation for government and industry's interlocking roles in strengthening the nation's R&D enterprise, our economy and ultimately, society. Thank you.

¹ S&E Indicators, 2004., p. 4-9, figure 4-2.
Return to speech.

² S&E Indicators, 2004., p. 4-13.
Return to speech.

³ Paraphrase taken from "Is Academic Science Driving a Surge in Industrial Innovation?", L. Branstetter, Columbia Business School, January 2003, p. 3.
Return to speech.