Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 5
Innovation in INFORMATION TECHNOLOGY 1 Innovation in Information Technology UNIVERSITIES, INDUSTRY, AND GOVERNMENT: A COMPLEX PARTNERSHIP YIELDING INNOVATION AND LEADERSHIP Figure 1 illustrates some of the many cases in which fundamental research in IT, conducted in industry and universities, led 10 to 15 years later to the introduction of entirely new product categories that became billion-dollar industries. It also illustrates the complex interplay between industry, universities, and government. The flow of ideas and people— the interaction between university research, industry research, and product development—is amply evident. Figure 1 updates Figure 4.1 from the 2002 CSTB report Information Technology Research, Innovation, and E-Government.1 The originally published figure2 produced an extraordinary response: it was used in presentations to Congress and to administration decision makers, and it was 1 Computer Science and Telecommunications Board, National Research Council. 2002. Information Technology Research, Innovation, and E-Government. National Academy Press, Washington, D.C. 2 Known informally as the “tire-tracks chart” because of its appearance, the figure was first published in Evolving the High Performance Computing and Communications Initiative to Support the Nation's Information Infrastructure (3; p. 2).
OCR for page 6
Innovation in INFORMATION TECHNOLOGY FIGURE 1 Examples of government-sponsored IT research and development in the creation of commercial products and industries. Federally sponsored research lies at the heart of many of today's multibillion-dollar information technology industries—industries that are transforming our lives and driving our economy. Ideas and people flow in complex patterns. The interaction of research ideas multiplies their effect. The result is that the United States is the world leader in this critical arena. Although the figure reflects input from many individuals at multiple points in time, ensuring readability required making judgments about the examples to present, which should be seen as illustrative rather than exhaustive. SOURCE: 2002 update by the Computer Science and Telecommunications Board of a figure (Figure ES.1) originally published in Computer Science and Telecommunications Board, National Research Council, 1995, Evolving the High Performance Computing and Communications Initiative to Support the Nation's Information Infrastructure, National Academy Press, Washington, D.C.
OCR for page 7
Innovation in INFORMATION TECHNOLOGY
OCR for page 8
Innovation in INFORMATION TECHNOLOGY discussed broadly in the research community. Although IT commercial success leads some policy makers to assume that industry is self-sufficient, the tire-tracks chart underscores how much industry builds on government-funded university research, sometimes through long incubation periods (1,3). Figure 1 also illustrates—although sketchily—the interdependencies of research advances in various subfields. There is a complex research ecology at work, in which concurrent advances in multiple subfields—in particular within computer science but extending into other fields, too, from electrical engineering to psychology—are mutually reinforcing: they stimulate and enable one another.3 One of the most important messages of Figure 1 is the long, unpredictable incubation period—requiring steady work and funding—between initial exploration and commercial deployment (1,3). Starting a project that requires considerable time often seems risky, but the payoff from successes justifies backing researchers who have vision. It is often not clear which aspect of an early-stage research project will be the most important; fundamental research produces a range of ideas, and later developers select from among them as needs emerge. Sometimes the utility of ideas is evident well after they have been generated. For example, some early work in artificial intelligence has achieved unanticipated applicability in computer games, some of which are now being investigated for decision support and other professional uses as well as recreation. It is important to remember that real-world requirements can change quickly. Although the end of the Cold War was interpreted by some as lessening the need for research,4 September 11, 2001, underscored research needs in several areas: system security and robustness, automatic natural language translation, data integration, image processing, and biosensors, among others —areas in which technical problems are difficult to begin with, and may become harder when technology must be designed to both meet homeland security needs and protect civil liber- 3 The idea that research in IT not only builds in part on research in physics, mathematics, electrical engineering, psychology, and other fields but also strongly influences them is consistent with what Donald Stokes has characterized in his four-part taxonomy as “Pasteur's Quadrant” research: use- or application-inspired basic research that pursues fundamental understanding (such as Louis Pasteur's research on the biological bases of fermentation and disease). See the discussion on pp. 26-29 in the 2000 CSTB report Making IT Better (2), and see Donald E. Stokes, 1997, Pasteur's Quadrant: Basic Science and Technological Innovation, Brookings Institution Press, Washington, D.C. 4 Linda R. Cohen and Roger G. Noll. 1994. “Privatizing Public Research,” Scientific American 271(3): 72-77.
OCR for page 9
Innovation in INFORMATION TECHNOLOGY ties.5 Without fundamental research, the cupboard is bare when there is a sudden need for ideas to reduce to practice. THE ESSENTIAL ROLE OF THE FEDERAL GOVERNMENT Federally sponsored research played a critical role in creating the enabling technologies for each of the billion-dollar market segments illustrated in Figure 1—and for many others as well. The government role coevolved with IT industries: its organization and emphases changed to focus on capabilities not ready for commercialization and on new needs that emerged as commercial capabilities grew, both moving targets (1). As this coevolution shows, successful technology development relies on flexibility in the conduct of research and in the structure of industry. Most often, this federal investment took the form of grants or contracts awarded to university researchers by the Defense Advanced Research Projects Agency (DARPA) and/or the National Science Foundation (NSF) —although a shifting mix of other funding agencies has been involved, reflecting changes in the missions of these agencies and their needs for IT (1,3). For example, the Department of Energy (DOE), the National Aeronautics and Space Administration (NASA), and the military services have supported high-performance computing, networking, human-computer interaction, and other kinds of research.6 Why has federal support been so effective in stimulating innovation in computing? As discussed below, many factors have been important. 1. Federally funded programs have supported long-term research into fundamental aspects of computing, whose widespread practical benefits typically take years to realize (1). “Long-term” research refers to a long time horizon for the research effort and for its impact to be realized. Examples of innovations that required long-term research include speech recognition, packet radio, computer graphics, and internetworking. In every case illustrated in Figure 1, the time from first concept to successful market is measured in 5 See Computer Science and Telecommunications Board, National Research Council. 2003. Information Technology for Counterterrorism: Immediate Actions and Future Possibilities. National Academies Press, Washington, D.C. 6 In addition to research funding, complementary activities have been undertaken by other agencies, such as the National Institute of Standards and Technology, which often brings together people from universities and industry on issues relating to standards setting and measurement.
OCR for page 10
Innovation in INFORMATION TECHNOLOGY decades (see Box 2)—a contrast to the more incremental innovations that are publicized as evidence of the rapid pace of IT innovation. Work on speech recognition, for example, which began in earnest in the early 1970s, took until 1997 to generate a successful product for enabling personal computers to recognize continuous speech (8). Work on packet radio also dates from the 1970s, and its realization in commercial ad hoc mobile networking also began in the late 1990s. 7 Fundamental algorithms for shading three-dimensional graphics images, which were developed with federal funding in the 1960s, saw limited use on high-performance machines until they entered consumer products in the 1990s; today these algorithms are used in a range of products in the health care, entertainment, and defense industries. The research programs behind these innovations not only were long-term but also were broad enough to accommodate within a single program the development of those unanticipated results that have in many cases provided the most significant outcomes of a project. The benefits of a long time horizon, combined with program breadth, extend to today's challenges. This point was emphasized in CSTB's 1997 report on usability, More Than Screen Deep (8), which explained (at p. 192): Federal initiatives that emphasize long-term goals beyond the horizon of most commercial efforts and that may thus entail added risk have the potential to move the whole information technology enterprise into new modes of thinking and to stimulate discovery of new technologies for the coming century. Because of unanticipated results and synergies, the exact course of fundamental research cannot be planned in advance, and its progress cannot be measured precisely in the short term. Even projects that appear to have failed or whose results do not seem to have immediate utility often make significant contributions to later technology development or achieve other objectives not originally envisioned. A striking example is the field of number theory (1): for hundreds of years a branch of pure mathematics without applications, it is now the basis for the public-key cryptography that underlies the security of electronic commerce. 7 Similarly, commercial developments in broadband cellular radio (which has become essentially wireless Internet access in third-generation wireless) are built in part on many decades of federally supported research into Code Division Multiple Access technology, signal processing for antenna arrays, error-correction coding, and so on.
OCR for page 11
Innovation in INFORMATION TECHNOLOGY BOX 2 The Role of Federal Support for Fundamental Research in IT CSTB's 1995 report Evolving the High Performance Computing and Communications Initiative to Support the Nation's Information Infrastructure (3) examined the payoff from several decades of federal investment in IT research. Among the conclusions of that report are these: Research has kept paying off over a long period. The payoff from research takes time. As Figure  shows, at least 10 years, more often 15, elapse between initial research on a major new idea and commercial success. This is still true in spite of today's shorter product cycles. Unexpected results are often the most important. Electronic mail and the “windows” interface are only two examples. . . . Research stimulates communication and interaction. Ideas flow back and forth between research programs and development efforts and between academia and industry [and between research programs with different foci that are proceeding concurrently]. Research trains people, who start companies or form a pool of trained personnel that existing companies can draw on to enter new markets quickly. Doing research involves taking risks. Not all public research programs have succeeded or led to clear outcomes even after many years. But the record of accomplishments suggests that government investment in computing and communications research has been highly productive.1 1 Computer Science and Telecommunications Board, National Research Council. 1995. Evolving the High Performance Computing and Communications Initiative to Support the Na-tion's Information Infrastructure. National Academy Press, Washington, D.C., pp. 3-4. 2. The interplay of government-funded and industry research has been an important factor in IT commercialization (1-8). The examples in Figure 1 show the interplay between government-funded research and industry research and development. In some cases, such as reduced-instruction-set computing (RISC) processors, the initial ideas came from industry, but the research that was essential to advancing these ideas came from government funding to universities. RISC was conceived at IBM, but it was not commercialized until DARPA funded additional research at the University of California at Berkeley and at Stanford University as part of its Very Large Scale Integrated Circuit (VLSI) program of the late 1970s and early 1980s (1,3). The VLSI program also supported university research that gave rise to such companies as Synopsys, Cadence, and Mentor, which have acquired dozens of smaller
OCR for page 12
Innovation in INFORMATION TECHNOLOGY companies that started as spinoffs of DARPA-funded8 university research; such research has also pushed the proverbial envelope in algorithms and user interfaces. The more than $3 billion electronic design automation industry is an essential enabler to other parts of IT. Similarly, IBM pioneered the concept of relational databases (its System R project) but did not commercialize the technology. NSF-sponsored research at the University of California at Berkeley brought this technology to the point at which it was commercialized by several start-up companies and then by more established database companies (including IBM) (1,3). In other cases, such as timesharing, the initial ideas came from the university community, and subsequent industry research, while significant for a time, was not sustained. In none of the examples in Figure 1 did industry alone provide the necessary research. 3. There is a complex interleaving of fundamental research and focused development (1-3). In the case of integrated circuit (VLSI) design tools, research innovation led to products and then to major industrial markets. A still-unfolding example is the theoretical research that yielded the algorithms behind the Web-content management technology underlying Akamai. In the case of relational databases, the introduction of products stimulated new fundamental research questions, leading to a new generation of products with capabilities vastly greater than those of their predecessors. The purpose of publicly funded research is to advance knowledge and to solve hard problems. The exploitation of that knowledge and those solutions in products is fundamentally important, but the form it takes is often unpredictable, as is the impact on future research (see Box 3). 4. Federal support for research has tended to complement, rather than preempt, industry investments in research. The IT sector invests an enormous amount each year in R&D. It is critical to understand, however, that the vast majority of corporate R&D has always been focused on product and process development (2). This is what shareholders (or other investors) demand. It is harder for corporations to justify funding long-term, fundamental research. Economists 8 In some cases, the Semiconductor Research Corporation provided the funding. For additional information, see the Web site <http://www.src.org/member/about/history.asp>. Accessed June 2, 2003.
OCR for page 13
Innovation in INFORMATION TECHNOLOGY BOX 3 The Technological Underpinnings of Electronic Commerce Electronic commerce is becoming pervasive. It is changing many aspects of our lives, from the way we shop to the way we obtain government services. The organizations and individuals that exploit electronic commerce employ commercial tools from companies such as Microsoft and Oracle. They may not think of themselves as the beneficiaries of federal investments in university-based IT research—but they are. Nearly every key technological component underlying electronic commerce has been shaped by this investment. For example: The Internet—Defense Advanced Research Projects Agency (DARPA) investments in the 1960s and 1970s were followed by National Science Foundation (NSF) investments in the 1980s and early 1990s, with research (supported by multiple agencies) continuing to this day (1). Web browsers—Mosaic, the first browser with a graphical user interface, was invented at the NSF-supported National Center for Supercomputer Applications at the University of Illinois (1). Public-key cryptography for secure credit card transactions—NSF sponsored university-based research in the 1970s that supported this innovation (1). Back-end database and transaction processing systems—NSF and DARPA supported key research on relational databases and transaction processing systems at the University of California at Berkeley, University of Wisconsin, and elsewhere, beginning in the early 1980s and continuing to this day (1). Search engines—Search engines grew out of federally supported university research programs, such as the ranking algorithm work at Stanford University that contributed to Google; the WebCrawler and MetaCrawler grew out of work at the University of Washington. But the development is not complete: a range of technical challenges still exist, along with challenges for improving the fit between the technologies and the behavior and needs of the people who use them (2,8). SOURCES: Pieces of this history are recounted in the previously cited CSTB reports (1-8) and in CSTB's series of reports on the Internet: Toward a National Research Network (1988), Realizing the Information Future: The Internet and Beyond (1990), The Unpredictable Certainty: Information Infrastructure Through 2000 (1996), The Internet's Coming of Age (2001), and Broadband: Bringing Home the Bits (2002), all published by the National Academy Press, Washington, D.C. have articulated the concept of “appropriability” to express the extent to which the results of an investment can be captured by the investor, as opposed to being available to all players in the market. The results of long-term, fundamental research are hard to appropriate for several reasons: they tend to be published openly and thus to become generally known; they tend to have broad value; the most important may be unpre-
OCR for page 14
Innovation in INFORMATION TECHNOLOGY dictable in advance; and they become known well ahead of the moment of realization as a product, so that many parties have the opportunity to incorporate the results into their thinking. In contrast, incremental research and product development can be performed in a way that is more appropriable: it can be done under wraps, and it can be moved into the marketplace more quickly and predictably. Although individual industrial players may find it hard to justify research that is weakly appropriable, it is the proper role of the federal government to support this sort of endeavor (1,3). When companies create successful new products using the ideas and workforce that result from federally sponsored research, they repay the nation handsomely in jobs, tax revenues, productivity increases, and world leadership (1,3). Long-term research often has great benefits for the IT sector as a whole, although no particular company can be sure of reaping most of these benefits. Appropriability helps to explain why the companies that have tended to provide the greatest support for fundamental research are large companies that enjoy dominant positions in their market (1). AT &T and IBM, for example, have historically made significant investments in fundamental research. Anything that advances IT as a whole benefits the dominant players—they may be capable of reaping a significant proportion of the returns on their research investments. As IT industries became more competitive, however, even these firms began to link their research more closely with corporate objectives and product development activities.9 One of them (AT&T) has radically cut back its research effort. This process began with a government proceeding that resulted in the splitting up of functions formerly aggregated under “Ma Bell” and continued with the growth and contraction of a set of industry research and development endeavors (AT&T Research, Lucent Technologies, Agere Systems, and Bellcore [now Telcordia]) where once there was the monolithic Bell Laboratories. 10 Several of the companies that have recently emerged as dominant in their sectors, such as Intel and Microsoft, have increased their support for fundamental research. However, many other successful companies with large market shares (e.g., Cisco, Dell, Oracle) have chosen not to invest in fundamental research to any significant extent. And even at Microsoft, just as at AT&T and IBM before it, the investment in fundamental research 9 Elizabeth Corcoran, 1994, “The Changing Role of U.S. Corporate Research Labs,” Re-search-Technology Management 37(4):14-20; Peter Coy, 1993, “R&D Scoreboard: In the Labs, the Fight to Spend Less, Get More,” Business Week, June 28, pp. 102-124. 10 CSTB launched a study of the future of telecommunications R&D in 2003.
OCR for page 15
Innovation in INFORMATION TECHNOLOGY represents a relatively small proportion of overall corporate R&D. In 2002, Microsoft invested roughly $5 billion in R&D, but the company's fundamental research arm is small enough to suggest that 95 percent of Microsoft's R&D investment is product-related. Start-ups represent the other end of the spectrum. A hallmark of U.S. entrepreneurship, start-ups and start-up financing promote flexibility in industry structure and industry management. They have facilitated the development of high-risk products as well as an iconoclastic, risk-taking attitude among more traditional companies and managers in the IT business. But they do not engage in research (2). Thus, the wave of Internet-related and other IT start-ups of the 1990s is notable for two reasons: first, these start-ups attracted some researchers away from universities and research, and second, notwithstanding the popular labeling of those start-ups as “high-tech, ” they applied the fruits of past research rather than generating more. Start-ups illustrate the critical role of government funding in building the foundations for innovative commercial investments. THE DISTINCTIVE CHARACTER OF FEDERALLY SUPPORTED RESEARCH The most important characteristic of successful government research activities is their breadth of scope—both in their long time dimension and in their focus on activities that are potentially difficult to appropriate privately in their entirety. Two specific topic areas that illustrate these principles are large-scale IT systems and social applications of IT. Growing capabilities and broadening use of IT in the 1990s motivated CSTB recommendations for greatly increased federal support in these two categories (2) (see Boxes 4 and 5). Prospects for progress in social applications—however difficult—are one reason for confidence that IT will improve as a human enabler. The beginnings evident in all of these areas are but crude indicators of what research may make possible. An example of particular currency is that of cybersecurity. Stimulated by the events of September 11, CSTB issued the report Cybersecurity Today and Tomorrow: Pay Now or Pay Later, in early 2002. The report summarized the findings of seven CSTB reports issued over the preceding decade that had cybersecurity as a principal theme. Cybersecurity Today and Tomorrow concludes with the following paragraph: Research and development on information systems security should be construed broadly to include R&D on defensive technology (including both underlying technologies and architectural issues), organizational and sociological dimensions of such security, forensic and recovery tools, and best policies and practices. Given the failure of the market to ad-
OCR for page 19
Innovation in INFORMATION TECHNOLOGY BOX 6 The Origins of Electronic Mail and Instant Messaging The invention of timesharing systems in the 1960s not only contributed important technical developments in hardware, software, and system security but also provided the environment that led to the development of the most useful and widespread of popular applications, namely, e-mail and instant messaging (1). Timesharing allowed concurrent multiple users to share the power of a computer, which provided a fresh way for colleagues to interact. By 1970, programmers in federally funded research laboratories had developed both asynchronous electronic mail and facilities for real-time interaction between users, in research operating systems such as Tenex, Multics, and CalTSS. These modalities—now widely known as e-mail and instant messaging— proved so powerful that they have spread far and wide with the availability of low-cost personal computers, public networking, and client-server computing. These popular and visible tools, as well as all of the other forms of collaborative computing, have truly transformed our work and our lives. They owe their origins to the funding of IT research by the Defense Advanced Research Projects Agency and the National Science Foundation (1,3). as specific technologies that can be translated into new open standards. 12 A 2002 report, Broadband: Bringing Home the Bits, outlines an even broader role for federally funded research to enable openness in infrastructural systems: Support research and development on access technologies, especially targeting the needs of nonincumbent players and other areas that are not targets of stable, private sector funding. . . . [One target area is] technologies that foster the accommodation of multiple competitive service providers over facilities. Such open access-ready systems might not be a natural research and development target of large incumbent providers but will be the preferred form for a variety of public sector or public-private deployments.13 Broadband: Bringing Home the Bits notes that federally funded research can complement the more proprietary-oriented industry approaches to innovation, whether in communications architecture or content. It also 12 Computer Science and Telecommunications Board, National Research Council. 2001. The Internet's Coming of Age. National Academy Press, Washington, D.C., p. 18. 13 Computer Science and Telecommunications Board, National Research Council. 2002. Broadband: Bringing Home the Bits. National Academy Press, Washington, D.C., p. 40.
OCR for page 20
Innovation in INFORMATION TECHNOLOGY calls for the support of research on economic, social, and regulatory factors relating to broadband technologies—nontechnical factors that interact with the design and deployment of broadband. UNIVERSITY RESEARCH AND INDUSTRIAL R&D Much of the government-funded research in IT has been carried out at universities.14 Federal support has constituted roughly 70 percent of total university research funding in computer science and electrical engineering since 1976 (2). Among the many benefits of federally funded university research, the generation of new knowledge is only one (see Box 7). Strong research institutions are recognized as being among the most critical success factors in high-tech economic development (5). In computing, electronics, telecommunications, and biotechnology, evidence of the correlation abounds—in Boston (Harvard University and the Massachusetts Institute of Technology); Research Triangle Park (Duke University, the University of North Carolina, and North Carolina State University); New Jersey (Princeton University, Rutgers University, and New York City-based Columbia University); Austin (the University of Texas); southern California (the University of California at San Diego, the University of California at Los Angeles, the California Institute of Technology, and the University of Southern California); northern California (the University of California at Berkeley, the University of California at San Francisco, and Stanford University); and Seattle (the University of Washington). In addition to creating ideas and companies, universities often import forefront technologies to their regions (e.g., the nationwide expansion of ARPANET in the 1970s and of NSFnet in the 1980s, and the continuation of those efforts through the private Internet2 activities in the 1990s and early 2000s). Universities also serve as powerful magnets for companies seeking to relocate. These contributions are not reflected in Figure 1. Figure 1 also does not capture the most important product of universities: people. The American research university is unique in the degree to which it integrates research with education—both undergraduate and graduate education. Not only do graduating students serve to staff industry (5,6), but they also are by far the most effective vehicle for technol- 14 The concentration of research in universities is particularly true for computer science research; industry played an important role in telecommunications research before the breakup of AT&T and the original Bell Labs.
OCR for page 21
Innovation in INFORMATION TECHNOLOGY BOX 7 The Diverse Benefits of University Research Universities have a number of important characteristics that contribute to their success as engines of innovation. Among them are the following: Universities can focus on long-term research. Focusing on long-term research is the special role of universities —one that IT companies cannot be expected to fill to any significant extent (1-3). America's IT companies are extraordinarily adept at improving current products, but the track record is at best mixed on the invention and adoption of “disruptive technologies,” and corporate research in IT has been becoming more applied (2). Universities provide a neutral ground for collaboration. Universities encourage movement and collaboration among faculty through leave and sabbatical policies that allow professors to visit industry, government, and other university departments or laboratories. These uniquely valuable components of the R&D structure in the United States are not generally present in industry. Universities also provide sites at which researchers from competing companies can come together to explore technical issues. At the same time that industry people share their wisdom and experience with university researchers, they have the opportunity to learn from one another (2,6). Universities integrate research and education. Universities provide a forum for educating the skilled IT workers of the future (5). The presence of research activities in an educational setting creates very powerful synergy (2,4). IT is a rapidly changing field. Many of the specific facts and techniques that a student learns become obsolete early in his or her career. The educational foundation for continuous learning—“keeping up with the field”—is a crucial component of IT education (5). Students, even beginning undergraduates, get that education not only in the classroom, but also by serving as apprentices on leading-edge research projects, where knowledge is being discovered, not read from a book. Often, new ideas are a by-product of what goes on in the classroom: in an attempt to explain the solutions to emerging problems, teachers often deepen their own understanding, while discovering interesting research questions whose answers are as yet unknown. Additionally, students are the most powerful vehicle for technology transfer, not only from university to industry but also between university laboratories and departments, through the hiring of postdoctoral researchers and assistant professors (5). Universities are inherently multidisciplinary. University researchers are well situated to draw on experts from a variety of other fields (2). There are often cultural barriers to cross-disciplinary collaboration, but physical proximity and collegial values go a long way in enabling collaboration. The multidisciplinary nature of universities is of historic and growing importance to computer science, which interfaces with so many other fields. Universities are “open.” This characteristic of universities, which is true both literally and figuratively, can pay enormous unanticipated dividends. Chance interactions in an open environment can change the world; for example, when Microsoft founders Paul Allen and Bill Gates were students at Seattle's Lakeside School in the early 1970s, they were exposed to computing and computer science at the University of Washington and a university spinoff company, Computer Center Corporation.
OCR for page 22
Innovation in INFORMATION TECHNOLOGY ogy transfer (see Box 7). Federal support for university research drives this process (1-6). In top university computer science programs, over half of all graduate students receive financial support from the federal government, mostly in the form of research assistantships. In addition, most of the funding for research equipment—that is, research infrastructure— comes from federal agencies. Industry also contributes significantly to equipment but is usually attracted by existing research excellence and collaborations. Thus, by placing infrastructure in universities, the federal government directly and indirectly makes possible hands-on learning experiences for countless young engineers and scientists, as well as enabling university researchers to continue their work (1-6). HALLMARKS OF FEDERALLY SPONSORED IT RESEARCH As discussed below, the hallmarks of federally sponsored IT research include scale, diversity, vision, and flexibility. 1. Federal programs have been effective in supporting the construction of large-scale systems and testbeds that have motivated research and demonstrated the feasibility of new technological approaches (1-3). Some research challenges are too large and require too much research infrastructure to be carried out by small, local research groups (6). In IT research, as in other areas of scientific investigation, federal programs have played an important role in stimulating and supporting large-scale efforts. DARPA's decision to construct a packet-switched network (called the ARPANET) to link computers at its many contractor sites prompted diverse, high-impact research on networking protocols, the design of packet switches and routers, software structures for managing large networks (such as the Domain Name System), and applications (such as remote log-in, file transfer, and ultimately the Web). Moreover, by constructing a successful system, DARPA demonstrated the value of large-scale packet-switched networks, motivating subsequent deployment of other networks—such as the NSF's NSFnet, which ultimately served as the foundation of the Internet—and also a series of high-speed networking testbeds (1,3). Much of the success of major system-building efforts derives from their ability to bring together large groups of researchers from universities and industry that develop a common vocabulary, share ideas, and create a critical mass of people who subsequently extend the technology (2,6).
OCR for page 23
Innovation in INFORMATION TECHNOLOGY 2. Computing research has benefited from diverse modes of research sponsored by different federal agencies (1-3). Funding for research in computing has been provided by various federal agencies—most notably DARPA and NSF, but also including other parts of the Department of Defense (DOD) besides DARPA, and other federal agencies such as NASA, DOE, and the National Institutes of Health (NIH; in particular through the National Library of Medicine). Complementary investments have supported technology transfer to industry (e.g., activities of the National Institute of Standards and Technology, or NIST). Funding agencies have continually evolved in order to match their structures better to the needs of the research and policy-making communities (1). (See Box 8.) In supporting research, these agencies pursue different objectives and employ different mechanisms. In contrast to NSF, for example —which has a mandate to support a very broad research agenda—“mission agencies” tend to focus on topics that appear to have the greatest relevance to their specific missions. Additionally, the early DARPA programs chose to concentrate large research awards in so-called centers of excellence (many of which over time have matured into some of the nation's leading university computer science programs), while NSF and the Office of Naval Research have supported individual researchers at a more diverse set of institutions (1). NSF has been active in supporting educational and research needs more broadly, awarding graduate student fellowships and providing funding for research equipment and infrastructure. CSTB has recognized the effective leadership of NSF and DARPA, calling on them to step up to larger roles (2; p. 11): The programs run by [NSF and DARPA] should complement one another and should together [do the following]: Support both theoretical and experimental work; Offer awards in a variety of sizes (small, medium, and large) to support individual investigators, small teams of researchers, and larger collaborations; Investigate a range of approaches to large-scale systems problems, such as improved software design methodologies, system architecture, reusable code, and biological and economic models . . . ; Attempt to address the full scope of large-scale systems issues, including scalability, heterogeneity, trustworthiness, flexibility, and predictability; and Give academic researchers some form of access to large-scale systems for studying and demonstrating new approaches.
OCR for page 24
Innovation in INFORMATION TECHNOLOGY BOX 8 Federal Agency Evolution In response to proposals by Vannevar Bush and others for an organization to fund basic research, especially in universities, the U.S. Congress established the National Science Foundation (NSF) in 1950 (1). A few years earlier, the U.S. Navy had founded the Office of Naval Research to draw on science and engineering resources in the universities. In the early 1950s, during an intense phase of the Cold War, the military services became the preeminent funders of computing and communications research. The Soviet Union's launching of Sputnik in 1957 raised fears in Congress and the country that the Soviets had forged ahead of the United States in advanced technology. In response, the U.S. Department of Defense, pressured by the Eisenhower administration, established the Advanced Research Projects Agency (ARPA, now DARPA) to fund technological projects with military implications. In 1962 DARPA created the Information Processing Techniques Office (IPTO), whose initial research agenda gave priority to further development of computers for command-and-control systems. With the passage of time, new organizations have emerged, and old ones have often been reformed or reinvented to respond to new national imperatives and counter bureaucratic trends (2). DARPA's IPTO has transformed itself several times to bring greater coherence to its research efforts and to respond to technological developments and changes in perceived national needs for IT. In 1967 NSF established the Office of Computing Activities, and in 1986 it formed the Computer and Information Science and Engineering Directorate to advance and coordinate support for research, education, and infrastructure in computing (1). In the 1980s NSF, which customarily has focused on fundamental research in universities, also began to encourage joint university-industry research centers through its Engineering Research Centers program (these centers focus on research and education in the context of long-time-horizon, complex engineering challenges 1) and its Science and Technology Center program (aimed at long-term research in areas that are new or that can bridge disciplines and/or institutions and sectors2). With the growth in the IT sector and corresponding IT development together with the maturation of the field of computer science, more recent federal funding has been characterized by a series of multiagency, long-term, high-risk initiatives. The first was the High Performance Computing and Communications Initiative, which emerged in the late 1980s and broadened through the mid-1990s (1,3). By the late 1990s and the establishment of the multiagency Information Technology for the Twenty-First Century initiative (in NSF, the Information Technology Research initiative), social science research—relating IT innovation to the people who use IT—was an important complement to the science and technology research per se (3,8). 1 See <http://www.eng.nsf.gov/eec/erc.htm>. Accessed June 2, 2003. 2 See <http://www.nsf.gov/od/oia/programs/stc/>. Accessed June 2, 2003.
OCR for page 25
Innovation in INFORMATION TECHNOLOGY Given the wide circle of agencies interested in and involved with IT research and the even wider circle coming to depend on large-scale IT systems, the NSF and DARPA should attempt to involve in their research other federal agencies . . . that operate large-scale IT systems and would benefit from advances in their design. Such involvement could provide a means for researchers to gain access to operational systems for analytical and experimental purposes. The diversity of research funding objectives and program management styles offers many benefits (1,3). It helps ensure exploration of a diverse set of research topics and consideration of a range of applications. For example, DARPA, NASA, and NIH (in addition to NSF) have all supported work in expert systems. However, because the systems have had different applications—decision aids for pilots, tools for determining the structure of molecules on other planets, and medical diagnostics— each agency has supported different groups of researchers who tried different approaches. And no one's judgment is infallible. If one agency declines to support a particular topic, researchers have other sources of funding. 3. Visionary program managers who were willing to take risks have been a hallmark of many of the highest-impact federal research initiatives (1,3). The program manager is responsible for initiating, funding, and overseeing research programs. The funding and management styles of program managers at DARPA during the 1960s and 1970s, for example, reflected an ability to marry visions for technological progress with strong technical expertise and an understanding of the uncertainties of the research process (1,3). Many of these program managers and program office directors were recruited from universities and industrial research laboratories for limited tours of duty and were themselves leading researchers. With close ties to the field, they were trusted by—and trusted—the research community. They tended to lay down broad guidelines for new research areas and to draw specific project proposals from principal investigators. They were willing to place bets—to pursue high-risk/high-gain projects. This style of funding and management allowed researchers room to pursue new venues of inquiry. The funding style resulted in advances in areas as diverse as computer graphics, artificial intelligence, networking, and computer architecture. As that experience illustrates, because unanticipated outcomes of research are so valuable, federal mechanisms for funding and managing research need to recognize the inherent uncertainties and build in enough flexibility to accommodate midcourse changes (1,3).
OCR for page 26
Innovation in INFORMATION TECHNOLOGY LOOKING FORWARD Federal funding agencies will have to continue to adjust their strategies and tactics as national needs and imperatives change. Today there is an escalation in concern about homeland security, the globalization of industry, a rise of commodity IT products and an IT mass market, the growing dependence of economic and social activity on networking and distributed computing capabilities, and a variety of industry retrenchments. Coevolution with industry thus means different things for federally funded computing research today than it did in the middle to late decades of the 20th century. Challenges as well as opportunities have grown: computer science is a larger field with more subdisciplines; telecommunications is increasingly intertwined with computing while evolving across multiple media;15 the interdisciplinary problems that engage computer science and telecommunications are broader-ranging; and the number of hard problems —reflecting growth in scale, complexity, and interactions with people —has increased. Evolving capabilities motivate a range of stretch goals that can help realize the potential of information technology as a human enabler.16 Examples include new forms of prosthetics (beginning with systems that can hear, speak, or see as well as a person can) and better ways to observe or participate in activities from a distance (i.e., telepresence). These circumstances imply that the challenge to federal research program managers has also grown. For example, while IT is at the core of a number of interdisciplinary programs (such as the multiagency Digital Libraries Initiative and NSF's Digital Government and Computing and Social System programs), it takes more work to review proposals for interdisciplinary work and to assure its quality. It may thus be more important to engage IT-using organizations in research projects, which may involve more work for the researchers (2). The growth in opportunities at the intersection of computing and biology, for example, or even computing and the arts—both topics of CSTB projects 17—suggests new horizons 15 Innovations are enhancing the potential of optical fiber, various forms of wireless, and even older media, such as copper. 16 These and other problems were outlined by Jim Gray in his 1998 A.M. Turing Award lecture. See Jim Gray. 1998. “What's Next? A Few Remaining Problems in Information Technology.” Available online at <http://research.Microsoft.com/~Gray/talks>. Accessed June 9, 2003. 17 The project on computing and the arts and design was completed in early 2003. See Computer Science and Telecommunications Board, National Research Council. 2003. Beyond Productivity: Information Technology, Innovation, and Creativity. National Academies Press, Washington, D.C.
OCR for page 27
Innovation in INFORMATION TECHNOLOGY for IT innovation that depend on the nurturing that is available through university-based research programs. The challenges confronting program managers underscore the need to attract talent from universities and industry to such public service positions. Past advances fostered by federal funding leveraged the energies and wisdom of people who went from universities and industry into the government, for at least a limited period. It is ironic that their success has increased the incentives for researchers to stay in universities or to try their hand in industry instead of cultivating the field as program managers. Government support for IT research will also be shaped by categories of problems in which it has a special interest. The events of September 11, 2001, remind us that computer and communications security, constrained by market failure, has always depended on federal investments. But so, too, has research in human-computer interaction, another arena in which market forces have been limited (8) and where the rise of e-government reinforces long-standing government interest associated with its own applications.18 The post-September 11 focus on homeland security and intelligence analysis also puts a spotlight on supercomputing architectures, numerical analysis, parallel programming languages and tools, and other areas in which IT advances have flowed from scientific and engineering computing needs within the research community at large—and in which purely commercial development was unlikely at best (1,3). The downturn in the telecommunications industry presents opportunities for the government to stimulate new directions through its support for research. We may see a consolidation and a loss of viable competition, or a realignment of the sector boundaries to better reflect economic realities. Government funding, supporting the development of open standards, can help shape the structure of industry.19 Given the “chicken-and-egg” tension shaping advances in infrastructure and applications, government support for exploration of new kinds of applications can have great impact.20 The government can encourage competition by supporting the definition of critical interfaces and demonstrations of feasibil- 18 Computer Science and Telecommunications Board, National Research Council. 2002. Information Technology Research, Innovation, and E-Government. National Academy Press, Washington, D.C. 19 See Computer Science and Telecommunications Board, National Research Council, 2001, The Internet's Coming of Age, National Academy Press, Washington, D.C.; and Computer Science and Telecommunications Board, National Research Council, 2002, Broadband: Bringing Home the Bits, National Academy Press, Washington, D.C. 20 This was demonstrated by the evolution of the early Internet and Web, involving development and refinement of both the underlying infrastructure and a suite of compelling applications by researchers focused not only on IT but also on other fields of science and engineering in which people used IT. The Internet probably could never have developed commercially without this phase of government-supported experimentation and refinement coordinated between infrastucture and applications. For a discussion of new opportunities in the support of applications, see Computer Science and Telecommunications Board, National Research Council, 2002, Broadband: Bringing Home the Bits, National Academy Press, Washington, D.C.
OCR for page 28
Innovation in INFORMATION TECHNOLOGY ity for open standards, and it can demonstrate new architectures through field trials and testbeds. This role was critical in the emergence of the Internet, and the relevance and importance of this sort of leadership have not waned.21 More generally, the 2001-2002 downturn in the economy and the crisis in the telecommunications industry caused a reduction in investment across all of IT. Spending remained down in 2003, and internal investment has dropped accordingly. Venture and equity capital has also become harder to obtain in the IT industries. In times such as these, research, especially longer-term research, is an obvious target for cost cutting. But if we as a nation do not continue to invest in the foundations of innovation, we run the risk that when an improving economy justifies an increase in investment, there may be few ideas in which to invest. For that reason this time is especially important for government-sponsored research. Today's research investments are essential to tomorrow's world leadership in IT. From its position of leadership today—reinforced by an aggregation of universities, companies, government programs, and talent—the United States is better positioned than other nations are to make the most of nonappropriable research (and even appropriable research). Properly managed, publicly funded research in IT will continue to create important new technologies and industries, some of them unimagined today. The process will continue to take 10 to 15 years from the inception of a new idea to the creation of a billion-dollar industry. Without continued federal investment in fundamental research there would still be innovation, but the quantity and range of new ideas for U.S. industry to draw from would be greatly diminished—as would the flow of people edu- 21 For a discussion of the role of government in setting a vision, see Computer Science and Telecommunications Board, National Research Council, 1994, Realizing the Information Future: The Internet and Beyond, National Academy Press, Washington, D.C. For a discussion of government leadership and the importance of government funding of research as a policy tool, see Computer Science and Telecommunications Board, National Research Council, 1996, The Unpredictable Certainty: Information Infrastructure Through 2000, National Academy Press, Washington, D.C.
OCR for page 29
Innovation in INFORMATION TECHNOLOGY cated at the forefront, the most important product of the nation's research universities (1-8). The lessons of history are clear, as many CSTB studies in the past decade have shown, and many of those lessons are relevant to 21st-century realities. A complex partnership among government, industry, and universities has made the United States the world leader in IT, and information technology has become essential to our national security and economic and social well-being. Turn-of-the-century turmoil and structural changes in IT industries have diminished their inherently limited capacity to support fundamental IT research. The role of the federal government in sponsoring fundamental research in IT— largely university-based—has been and will continue to be essential.
Representative terms from entire chapter: