Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Executive Summary Universities have a long history of contributing to industry and the economy in a variety of ways. Yet only recently has attention been focused systematically on the nature and extent of these contributions to specific industries. With sup- port from the Alfred P. Sloan Foundation, the National Academy of Engineering initiated a study in 1998 to document and assess the contributions of academic research to the performance of five industry sectors: network systems and com- munications; medical devices and equipment; aerospace; transportation, distribu- tion, and logistics services; and financial services. A 15-member committee of industry and university experts conducted the study, with integral support from five panels (one for each industry). Since this study began, there has been growing recognition of the importance of universities and academic research to industrial innovation and performance. There has also been greater recognition of how the role of academic research has changed over time, especially over the last 25 years as many universities have become more directly involved in the commercialization of the fruits of their research. This study documents the contributions of university research to five very different industry sectors and provides a qualitative assessment of the im- pact of that research. The committee's conclusions cut across all five sectors; the recommendations propose steps that should be taken to meet a variety of chal- lenges in academic research and industry-university interactions. Few data are available to support a quantitative assessment of the impact of academic research on the performance of specific industries. Quantitative mea- sures of the output of academic research seldom go beyond counts of research papers, patents, and royalty income, none of which directly correlates with the impact on industry. Industrial performance can be measured by shareholder
2 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE value, employment growth, market share, technical advances, and other factors, all of which tend to be cyclical and company specific. Therefore, the committee relied on informed opinion and expert judgments, supplemented by reviews of the literature, e-mail surveys, workshop discussions, and panel deliberations, to make qualitative assessments of the impact of academic research. All five industry sectors examined in this study are important in terms of sales and employment, technological intensity, and expected growth rates. Two sectors, medical devices and equipment and network systems and communica- tions, have a recent history of extensive collaboration with academic researchers. The other three, aerospace, financial services, and transportation, distribution, and logistics services, have had more limited, less systematic interactions with the academic research enterprise. All five have demonstrated significant capaci- ties for innovation over the past decade. However, the financial services industry and the transportation, distribution, and logistics services industry, the two pre- dominantly service industries, have only recently begun to integrate research and development (R&D) into their way of doing business. Ultimately, the committee concluded that academic research has had a significant impact on performance in the network systems and communications, medical devices and equipment, and financial services industries. The impact on performance in the aerospace and the transportation, distribution, and logistics services industries has been moderate. Academic research has made substantial contributions to all five industries, ranging from graduates at all levels trained in modern research techniques to fundamental concepts and key ideas based on basic and applied research to the development of tools, prototypes, and marketable products, processes, and ser- vices. The disciplinary sources of research contributions span the fields of engi- neering, the natural sciences, and the social and behavioral sciences, with ad- vances in one field often combining and building on developments in other fields. Pathways linking academic research and industries include: direct hires of stu- dents, graduates, and faculty; temporary exchanges of researchers; faculty consultancies; joint research involving industry and academic scientists and engi- neers; industry-sponsored research contracts and grants; a variety of institutional mechanisms at universities (e.g., research centers, consortia, industrial liaison programs); technology licensing; start-up companies; publications; conferences; and short courses. Each industry studied illustrates a distinct pattern of collaboration with universities and has developed different mechanisms for taking advantage of academic contributions. For example, network systems has a history of drawing upon academic research for fundamental innovations, as well as using universi- ties as test beds for new networking concepts that have provided the under- pinnings of the Internet, the World Wide Web, and e-commerce. Figure ES-1 shows the complex pattern of academic and industrial research over many years in this industry. The medical devices and equipment industry has also looked to universities for fundamental multidisciplinary research in the physical sciences
EXECUTIVE SUMMARY 3 and engineering and the unique capabilities of academic medical centers for researching, developing, testing, and improving devices, as well as conducting the clinical trials necessary to obtaining regulatory approval, all in an atmo- sphere of close industry-university collaboration. In financial services, contri- butions of academic research in economics, engineering, and mathematics have been important to the development of new financial models and instruments, in spite of the industry's lack of a well developed R&D infrastructure. This study confirms that graduates trained in research are a major component of academia's contribution to industrial performance. U.S. universities are indus- tries' primary source of people with research training and experience under- graduates, graduate students, postdoctoral researchers, and faculty. Research- trained students and graduates at all degree levels play a critical role in the development, transfer, diffusion, and application of new knowledge and technol- ogy both within and between industry and academia. A second consistent message that emerges from a study of these five indus- tries is that basic, long-term research performed by universities across a wide range of science and engineering disciplines has made significant contributions to industry over time. For instance, portfolio theory, linear programming, deriva- tive pricing theory, and prospect theory, all of academic origin, have laid the foundation for whole new families of financial products and services. Academic contributions to linear and integer programming and to queue theory are the building blocks of the information-management and decision-support technolo- gies at the heart of the integrated logistics revolution. Medical devices, such as magnetic resonance imaging machines and pacemakers, are based on the contri- butions of fundamental research arising from multiple disciplines in the natural . . . sciences ant englneenng. Universities perform roughly half of all basic research in the United States, most of it funded by the federal government. Over time, basic research results build on one another and intermingle with results in other fields, often through the free exchange of people and ideas that universities facilitate. Typically pur- sued with no specific commercial application in mind, basic research has pro- vided the technological underpinnings for commercial innovation. Relevant re- search findings have come from multiple disciplines, indicating that a portfolio of research investment in many different fields is essential to continued progress. In some cases, the original academic source of a basic understanding of a process or technology is forgotten by the time the knowledge is put to use; in other cases, Nobel prizes are awarded years later for fundamental breakthroughs. Individual academic researchers may or may not receive the credit they deserve for the constant flow of commercial innovations, but there can be no doubt that basic research and the publication of its results have made a unique and essential contribution to industrial performance. Contributions from applied academic research are also very important to industry. Today, academic applied researchers and their academic research
4 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE 1965 1970 1975 1980 1985 1990 1995 2005 Timesharing ClienVserver imputing Graphics ~ _ NIP. Internet LANs. . VLSI design RISC processors Workstations ~ 1 an I ~ Graphical user interfaces . I ~ ~ 1 to World Wide Web I 1965 1970 1975 1980 1985 1990 1995 2005 | University ~ Industry R&D CTSS, Multics / BED Unix SDS 940, 360/67, VMS Berkeley, CMU, CERN PARC, DEC, IBM Novell, EMC, Sun, Oracle Sketchpad, Utah GM/IBM, Xerox, Microsoft E&S, SGI, ATI, Adobe Spacewar (MIT), Trek (Rochester Atari, Nintendo, SGI, Pixar ARPANET, Aloha, Internet Pup DECnet, TCP/IP Rings, Hubnet Ethernet, Datakit, Autonet LANs, switched Ethernet Lisp machine, Stanford Xerox Alto Xerox Star, Apollo, Sun Engelbart / Rochester Alto, Smalltalk Star, Mac, Microsoft Berkeley, Caltech, MOSIS many Berkeley, Stanford IBM 801 SUN, SGI, IBM, HP · - - - - - Products ~ $1 bit. market | FIGURE ES-1 Examples of academic government-sponsored (and some industry- sponsored) IT research and development in the creation of commercial products and industries. Source: NRC, 2003.
EXECUTIVE SUMMARY 1965 1970 1975 1980 1985 1990 1995 2005 Relational databases ~from/Tternet Parallel databases Datarnining Parallel computing RAIDjdisk servers Portable communication I.... 1 ~2 \-- . . . World Wide Web _, . . , . . , Speech recognition Broadband in last mile 1965 1970 1975 Berkeley, Wisconsin IBM Oracle, IBM, Sybase Tokyo, Wisconsin, UCLA IBM, ICE JCL, Teradata, Tandem Wisconsin, Stanford IBM, Arbor IRI, Arbor, Plato Illiac 4, CMU, Caltech, HPC IBM, Intel CM-5, Teradata, C ray T3D Berkeley Striping/Datamesh, Petal many Berkeley, Purdue (COMA) Linkabit, Hughes Qualcomm ~ ~ ~ , \\ · ~ ~ ~ ~ - 1980 1985 1990 1995 2005 CERN, Illinois (Mosaic) Alta Vista Netscape, Yahoo, Google CMU, SRI, MIT Bell, IBM, Dragon Dragon, IBM Stanford, UCLA Bellcore (Telcordia) Amati, Alcatel, Broadcom | University s ~ Industry R&D · ~ ~ - - - Products bit. market |
6 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE infrastructure are directly involved in the development of industrial tools, proto- types, products, and production processes, as well as the delivery of products and services. Box ES-1 shows examples of specific applied academic research contributions to each of the five industries. Sometimes academic applied re- search yields cumulative, incremental advances over time that prove to be of major importance to industry. Individual companies have also greatly benefited from university-based research to solve discrete practical problems related to
EXECUTIVE SUMMARY 7 their businesses. University-based research centers, with industrial participa- tion, have become another important venue for applied research as well as more directed basic research of value to industry. Applied research through multidisciplinary collaboration among science, engineering, and/or medical fac- ulties is a unique strength of academia. Most of the funding for applied research at universities comes from federal agencies that want specific problems solved related to their missions. Industry funds some academic applied research di- rectly via research contracts and also indirectly through support of university- based research centers. Industry funding of academic research increased throughout the 1990s, but industry still provides only a small proportion of total funds for academic research. The natural sciences and engineering disciplines are not the only fields that make important research contributions to industrial innovation and performance. The contributions of the social and behavioral sciences to industry have been greatly undervalued. Researchers in business, economics, psychology, and many other fields have made valuable contributions to progress in the industries stud- ied. For example, in network systems and communications, social and behav- ioral scientists have provided the knowledge base for the formulation of regula- tory policy and have generated useful knowledge and insights into group dynamics and decision making, the diffusion of new technologies, the nature and value of network externalities, and the relationship between organizational char- acteristics and information dissemination and sharing. CHALLENGES AND OPPORTUNITIES The industry studies revealed six major, crosscutting challenges and oppor- tunities for university-industry research that warrant careful attention by univer- sities, industry, and government. First, there is a growing imbalance in federal R&D funding. Current invest- ments in life sciences far outpace investments in the complementary disciplines of physical sciences, engineering, and the social and behavioral sciences. The industry studies show time and again that the value of research results in one field often depends heavily on advances in complementary fields. In addition, univer- sities must maintain a mix of basic and applied research to sustain their role as repositories of expertise and resources in many disciplines. Federal funding is now virtually the only source of support for basic research, which makes the effective management of federal research programs of paramount importance. Program managers in federal agencies must work with academic and industrial researchers to develop research agendas that might lead to major new insights. The challenge is not just to maintain a balance between basic and applied re- search, but also to ensure that the basic research portfolio is sufficiently diverse to stimulate innovative thinking by academic researchers in many fields.
THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE Second, the industry studies underscore the need for industries and universi- ties to continue exploring mechanisms and pathways for bringing the benefits of academic research to industry, keeping in mind that what works well in one industry may not work well in another. Both partners should experiment with new approaches. University-industry linkages must be adaptable, and universi- ties should be on the lookout for opportunities to link up with new industries and explore leading-edge industrial research activities and challenges. Cross-sectoral movement and interaction between individual academic and industrial research- ers are essential to promoting the effective two-way exchange of knowledge and technology. Third, the studies suggest that services industries represent a significant source of opportunity for university-industry interaction. Services account for more than 80 percent of the U.S. gross domestic product, employ a large and growing share of the science and engineering workforce, and are the primary users of information technology. In most manufacturing industries, service func- tions (such as logistics, distribution, and customer service) are now leading areas of competitive advantage. Innovation and increased productivity in the services infrastructure (e.g., finance, transportation, communication, health care) have an enormous impact on productivity and performance in all other segments of the economy. Nevertheless, the academic research enterprise has not focused on or been organized to meet the needs of service businesses. Major challenges to services industries that could be taken up by universities include: (1) the adapta- tion and application of systems and industrial engineering concepts, methodolo- gies, and quality-control processes to service functions and businesses; (2) the integration of technological research and social science, management, and policy research; and the (3) the education and training of engineering and science gradu- ates prepared to deal with management, policy, and social issues. Fourth, in recent decades, universities have increasingly emphasized tech- nology transfer and the generation of income from research activities by patent- ing and licensing research results and the creation of technology transfer offices. The increased attention to the management of intellectual property has had many positive consequences for both industry and academia, including providing in- centives for invention and innovation and the dissemination and commercializa- tion of new technologies. However, many questions remain regarding the net returns on investments in technology transfer, as well as their impact on the core research and educational missions of universities. These are questions that aca- demic researchers are well equipped to address. Fifth, regulation and regulatory changes continue to have a profound influ- ence on industry receptivity to contributions by academic research. In turn, academic research continues to shape the regulatory environment of many in- dustries, including most of those in this study. Research universities are well equipped to provide interdisciplinary expertise that can inform ongoing regula- tory debates in these and other industries. Moreover, the influx of academically
EXECUTIVE SUMMARY 9 trained scientists and engineers into financial, health, communication, transpor- tation, and other regulatory bodies has strengthened these agencies' ability to draw effectively on research advances across a spectrum of disciplines relevant to their tasks. The rapid pace of technological change in many sectors and its effect on the structure, competitive dynamics, and economic and societal im- pacts of many industries underscores the importance of sustaining, as well as deepening, the current productive relationship between academic researchers and regulators. Finally, information technologies are critical to the performance of all indus- tries and will continue to be so in the future. Industry's need for the continued development, diffusion, and effective application of advanced information tech- nologies presents major opportunities for academic research in mathematics, computer sciences, physical sciences, life sciences, multiple engineering disci- plines, and social and behavioral sciences. CONCLUSION The committee's review of these five very different industries shows that academic research has clearly provided benefits to industry and has had a posi- tive, long-term impact on industrial performance. However, it is difficult to identify specific mechanisms by which this impact can be maximized. The research competencies, the ability to interface with industry, the quality of infra- structure, and many, many other characteristics vary greatly from university to university. In addition, industries in the abstract do not interface with universi- ties; only individual companies do. And companies in a given industry also vary in their ability to manage interfaces with universities, in their expectations of what academic researchers can provide, in the complexity of their research prob- lems, and in their time horizons. And all of these vary over time. When this study began, high-technology industries, such as network systems and communi- cations, were booming, attracting academic researchers and potential graduate students to well funded industrial laboratories, growing operations, and a plethora of start-up companies. As the study comes to a close, this same industry is experiencing decreases in sales, stock prices, investments, research funding, and employment. Thus, both the unique characteristics of individual institutions and the changes brought about by economic cycles must be kept in mind in assessing the impact of academic research. Ultimately, the study of these five industries underscores the core strengths of the academic research enterprise. Because universities are venues for a greater range of ideas and disciplinary perspectives than any other institutions in the U.S. innovation system, they have vast potential for multidisciplinary re- search. In addition, universities are the only places where advanced research and education are integrated on a large scale. The constant flow of new students through universities continuously revitalizes the academic research enterprise,
10 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE challenging the assumptions of faculty and bringing fresh perspectives to re- search. Indeed, the rich and varied interaction between university research faculty and students and companies in many industries exposes the former to industrial challenges that often serve as stimuli to basic and applied research in academia. These core strengths augur well for the future of academic research and its continuing contributions to industry performance. GENERAL RECOMMENDATIONS The general recommendations of this study call for actions that could en- hance the contributions of university research to industrial growth and perfor- mance. Recommendations for individual industries can be found in the body of the report. The general recommendations address crosscutting challenges and opportunities that apply to more than one industry. General Recommendation 1. Because the contributions of academic research are diverse and often indirect, a broad and balanced portfolio of academic re- search should be maintained. Recent trends in federal funding indicate that funding levels for research in the physical sciences, engineering, and the social and behavioral sciences should be increased. . . Congress and the administration should restore the balance in federal funding of academic research by increasing support for research in the physical sciences, engineering, and the social and behavioral sciences to complement and leverage the results of recent heavy investments in the life sciences and medical sciences. Federal funding of academic research should continue to emphasize long- term basic research, as well as applied research (typically funded by mission agencies). Multidisciplinary research should be encouraged through support of project-specific research teams and other institution- alized mechanisms, such as engineering research centers and other university-industry research centers. General Recommendation 2. Industries and universities should continue to explore mechanisms and pathways for bringing the benefits of academic re- search to industry, keeping in mind that what works well in one industry may not work well in another. Both partners should experiment with new approaches. University-industry research linkages should be adaptable, and universities should be on the lookout for opportunities to link up with new industries and explore leading-edge industrial research activities and challenges. Given the importance of personal relationships among academic and indus- trial researchers for productive collaboration and knowledge transfer, universities
EXECUTIVE SUMMARY 11 and industry should foster interactions between university- and industry-based scientists and engineers in the following ways: · A major program of fellowships should be established to attract and sup- port graduate students in science and engineering. · Sabbatical programs should be established and/or expanded to encourage academic and industrial researchers to spend time in each other's home research settings. More balanced participation by academic researchers and their industry counterparts in major conferences on specific sectors, technical systems, and disciplines should be encouraged. · New ways of supporting personal interactions across academia-industry boundaries, including using technology to support collaboration, should be explored. University-industry research centers should be structured to facilitate close interaction between academic and industrial researchers. Academic reward structures, such as promotion and tenure criteria, should be reviewed and modified (as necessary) to encourage and reward re- searchers who attract research support from industry and/or address sig- nificant research questions of direct importance to industry. Intellectual property rights policies and practices that facilitate productive research collaboration with industry should be promulgated at universities. . General Recommendation 3. The ability of academic researchers to contribute to services industries and the receptivity of leaders in the services industries to the potential contributions of academic research must both be improved. The following steps would have immediate benefits: . Academic research contributions and capabilities relevant to each indus- try should be documented and promoted in the targeted communities to educate senior managers about how academic research might improve company performance in the marketplace. · Common legal frameworks acceptable to industry and academia should be established detailing the terms of confidentiality and related conditions to facilitate academic researchers' access to operational networks and real-time data. · Federal mission and regulatory agencies with primary responsibility for the services industries (e.g., Securities and Exchange Commission, Inter- nal Revenue Service, Federal Communications Commission, and U.S. Department of Health and Human Services) should consider funding aca- demic research in ways that encourage greater participation by the ser- vices industries. Engineering research centers funded by the National
12 THE IMPACT OF ACADEMIC RESEARCH ON INDUSTRIAL PERFORMANCE Science Foundation and university transportation centers funded by the U.S. Department of Transportation could serve as models. General Recommendation 4. Individual researchers and organizations, such as the Association of University Technology Managers, that gather data on univer- sity research and technology-transfer activities should continue to monitor and assess the effectiveness of incentives for transferring academic research results (particularly intellectual property policies and practices) and the impact of entre- preneurial activity by academic researchers on the traditional university missions of education, research, and service. The following issues should be addressed: . The costs to institutions of patenting research results, including the costs of maintaining and defending patents, should be assessed and compared to the benefits, in terms of income from licenses and royalties. Steps being taken to disseminate patent information to improve the chances of commercialization should be reviewed and best prac- tices identified. Best practices in the long-term management of patent inventories should be shared among research institutions. · The effectiveness of technology transfer via patented inventions should be assessed and compared to transfer via more traditional mechanisms, such as publications. The benefits to faculty and universities should also be compared. The impact of university-industry research collaboration and technology transfer activities on undergraduate, graduate, and continuing education, the composition of academic research, the stability of academic research funding, the private and social returns from academic research, the many traditional service roles of the university, and other related issues should be assessed. General Recommendation 5. Government regulatory agencies, including the Food and Drug Administration, the Environmental Protection Agency, the Fed- eral Communications Commission, and the Securities and Exchange Commis- sion, should be encouraged to sustain and strengthen their productive interaction with academic researchers and to continue to explore new mechanisms for bring- ing scientific and engineering advances, including scientifically based concepts and tools, to bear more rapidly and effectively on regulatory processes. General Recommendation 6. Government, industry, and universities should work together to meet the challenges and opportunities created by information technologies. The following steps would be beneficial:
EXECUTIVE SUMMARY 13 Boost federal funding for fundamental research in information technolo- gies, as part of an effort to redress the imbalance in federal funding for various disciplines in academic research. Increase public and private sector investment in software research, with an emphasis on (1) engineering methods for assessing and improving quality and (2) software that is more flexible and responsive to changing business conditions. Support more interdisciplinary research on existing and potential infor- mation technologies that combines engineering methods and the social and behavioral sciences. REFERENCE NRC (National Research Council). 2003. Innovation and Information Technology. Washington, D.C.: The National Academies Press.