NRC Says More Research on Information Technology Necessary

Washington, D.C. -- Government, academia and business should all make a greater commitment to research on the potential and inherent problems of information technology (IT), according to a report released May 10 by the National Research Council (NRC) entitled "Making IT Better: Expanding Information Technology Research to Meet Society's Needs."

The co-chair of the authoring committee presented the findings in a Washington briefing, and acknowledged that IT research is drawing more investment than ever. But still, David Messerschmitt asked, "Is funding for IT research sufficient to keep up with the growing challenges it presents? Are the right kinds of research being supported? And are the right structures in place for funding and conducting that research? Our report says 'no' to each of these questions."

The report recommends:

• The federal government should continue to increase funding for fundamental IT research.
• The National Science Foundation and the Defense Advanced Research Projects Agency should establish programs of fundamental research in the area of large-scale systems.
• Federal agencies should support more interdisciplinary work related to social applications of information technology.
• Universities should encourage additional research across disciplines to explore uses and applications of information technology.
• IT corporations should become more involved in research.
• IT corporations should actively engage users and customers in their research efforts.

More information on the report is available at http://www.nationalacademies.org/

Following is the transcript of the opening statement:

(begin transcript)

Making IT Better: Expanding Information Technology Research to Meet Society's Needs

National Research Council

Public Briefing

May 10, 2000

Opening Statement by David G. Messerschmitt, Co-Chair, Committee on Information Technology Research in a Competitive World

Good morning, and welcome to the National Academies. It is my privilege to present to you the main conclusions of the report Making IT Better: Expanding the Scope of Information Technology Research to Meet Society's Needs, which the National Research Council is releasing today.

As our committee found in its study, information technology research in the United States is, in many respects, incredibly healthy. Since 1995, both industry and government have steadily increased their investments in IT research. IT companies invested almost $45 billion in research and development during 1998 -- $13 billion of which was classified as research. Government agencies invested another $1.4 billion in IT-related research that same year. Proposals are in place to boost that figure by $1 billion over several years.

Although more money is being spent on research than ever before, a number of questions remain. Is funding for IT research sufficient to keep up with the growing challenges it presents? Are the right kinds of research being supported? And are the right structures in place for funding and conducting that research? Our report says "no" to each of these questions.

Over the past several decades, information technology has moved out of the laboratory and the back rooms of large organizations and now touches virtually all aspects of life. The pace of this transition has quickened in recent years due largely to the Internet. Research has created the knowledge, insight, and understanding that have allowed innovators to develop more powerful computing and communications systems that have fueled these developments.

As information technology itself has changed, so too does research need to change. Today's research continues, in large part, to pursue further improvements in the components that make up larger IT systems. But this research has only just begun to achieve the understanding needed to develop large-scale systems that are reliable, secure, easy to use, and easy to modify. Our report thus advocates an expansion of research into this area.

Evidence of the limitations in our understanding of large-scale systems is all around us. By some estimates, as many as 70 percent of all IT system development efforts fail to meet their requirements, are cancelled, or exceed their cost and time budgets. The difficulties encountered by the Internal Revenue Service, Federal Aviation Administration, and Bureau of Land Management in modernizing their information systems are among the more widely publicized failures, but government is not alone in facing these challenges. Many private sector organizations also have experienced extended system outages or difficulties implementing new information technology systems.

A major reason for these failures is inadequate understanding and a lack of suitable design methodologies for large-scale systems. Today's IT systems are larger, more heterogeneous, and more complex than their predecessors. They have billions, or even trillions, of different components that interact with each other in complicated and often unexpected ways. Over time, systems of a certain scale and complexity are not feasible or economical to design with existing methodologies. Research is needed to define better organizational paradigms and design methodologies to extend the range of the feasible and economic far beyond what is considered possible today. Our hope is that such research will result in a solid foundation of generic approaches for designing large-scale systems as well as more specific approaches that work on particular systems.

Problems with large-scale systems are not new to computing. They have long been recognized by the research community. But to date, they have received a level of attention from researchers that pales in relation to the challenges and opportunities they present. These problems will become more limiting in the future as the size of IT systems grows and society becomes more dependent on them for everything from government, banking, and manufacturing to health care and education. Continued failures in system development and performance cannot be tolerated. It is time that we learned more about how to design large systems properly.

The problems that need to be addressed with IT systems are not purely technical in nature. Information technology is being increasingly deployed in support of what we call "social applications" that serve groups of people in shared activities. The most straightforward of these applications improve the effectiveness of geographically dispersed groups of people who are collaborating on some task in a shared context. More sophisticated applications may support the operations of a business or the functioning of an entire economy; systems for e-commerce are an example. Social applications embed information technology into a large organizational or social system to form a "sociotechnical system." That is, a system in which people and technology interact to achieve a common purpose. As a result, social applications mix technical and nontechnical design and operational elements and involve often-difficult social and policy issues such as those related to privacy and limited access.

Research is needed to address such social applications. This research will be interdisciplinary by its very nature. It should involve not only technologists, but researchers from the social sciences, business, and law, as well as users of IT systems in particular application domains. The interaction among users, technologists, and scientists can influence the direction of research and make the outcomes more immediately beneficial. There is an opportunity to redefine the technology to better fit into and support the organizations and the processes it is supporting. For example, collaboration on health care information systems could uncover the need for particular kinds of security and privacy features that might otherwise be overlooked.

Today's mechanisms for funding and conducting research are not well-suited to work in large-scale systems and social applications of information technology. In industry, for example, most of the companies that conduct research tend to focus their attention on particular components of IT systems, whether microprocessors, computers, Internet routers, or software, rather than on large-scale systems or applications. Those companies that specialize in bringing together large IT systems -- the so-called systems integrators and enterprise and commerce applications suppliers -- tend not to conduct research at all. Neither do the organizations that operate and use these systems. Despite a long history of research in the telecommunications services industry, most network operators today conduct little or no research. Instead they rely on their equipment suppliers. But these suppliers tend to focus on specific components like switches and fiber optics rather than on systems as a whole or social applications of these systems.

Academic researchers are also limited in addressing system issues. Many of the problems observed with IT systems are manifest only in large-scale operational systems to which researchers have limited or no access. The systems and the problems they exhibit are difficult to model and simulate in a laboratory setting. Social applications of IT systems are even more difficult to examine since they require close contact with expert users or researchers in disciplines other than information technology.

New mechanisms are needed for funding and conducting research. These mechanisms should be designed specifically to address problems of large-scale systems and social applications of information technology and should attempt to better link the traditional IT research community to users and to researchers in complementary disciplines. This work will necessarily require some degree of collaboration among groups of researchers and funding levels capable of sustaining such interaction over longer periods of time. This is not to say that individual investigator research or work on IT components is no longer needed. Just the opposite -- continued efforts are needed in these areas to ensure that progress is made and that revolutionary new ideas in computing and communications are pursued. But additional effort is needed to ensure that work on large-scale systems and social applications of information technology can thrive. Doing so will require the combined efforts of government, industry, and universities.

The committee recommends several steps that these partners should take:

First, the federal government should continue to boost funding for fundamental IT research. Additional funding is necessary to ensure that continued progress is made in basic computing and communications technologies, and that the research agenda expands more fully into the areas of large-scale systems and social applications. Industry cannot be expected to support this new work to a suitable degree because it is highly risky and may not produce tangible benefits in the near term. Just how much additional funding is needed is something that will become evident over time as researchers rally around the larger set of problems. The committee believes that the increases that have already been projected for information technology -- on the order of hundreds of millions of dollars per year -- are of the right order of magnitude.

Second, the National Science Foundation (NSF) and Defense Advanced Research Projects Agency (DARPA) should establish programs of fundamental research in the area of large-scale systems. These agencies already have several projects under way in this area, but a more systematic and comprehensive effort is needed that supports both theoretical and experimental work and examines a range of potential solutions, from new architectures and design principles, to analogies from the biological and economic sciences. This research should engage IT experts from other federal agencies and industry that make use of large-scale IT systems. This will allow researchers to gain greater insight into the kinds of problems these systems pose and into the feasibility of different solutions.

Third, federal agencies should support more interdisciplinary work related to social applications of information technology. These efforts could build on existing programs, such as the digital government initiative and the computing and social sciences program at NSF. They should also include larger grants that can support the work of small interdisciplinary research teams and of participants in larger research centers that bring together teams for extended periods of time. NSF already plans to support research centers on the social and economic impacts of information technologies. These centers should be designed to also consider the implications of social applications on IT research. While it makes sense for agencies with experience in funding IT research, such as DARPA or NSF, to lead such efforts, they should explicitly require the participation of other federal agencies and, where possible, of institutional users in the private sector. The greatest challenge will be in ensuring suitable mechanisms for reviewing proposals and monitoring the quality of the research conducted. Review teams should have a breadth of expertise as expansive as that involved in the research and as varied as the many research proposals that are funded.

Fourth, universities should take steps to encourage and facilitate additional interdisciplinary research and education related to information technology. Additional federal funding will provide a strong incentive for researchers from different disciplines to work together on social applications of information technology, but faculty may be discouraged from participating in such efforts by tenure and review processes that are strongly rooted in disciplinary perspectives. New interdisciplinary programs, schools, or departments with their own hiring and promotion authority are one possible solution that has been pursued by a handful of universities. Another option is to ensure that existing academic departments develop criteria for evaluating work conducted across disciplinary boundaries. Senior faculty will undoubtedly have to lead the way in these changes.

Fifth, corporations that are significant end users of Information Technology or that integrate IT systems should become more actively engaged in research. This does not mean that these organizations should set up internal research labs and start funding university researchers -- at least not immediately. Most lack the experience to effectively conduct and manage such work. But end-user organizations can gain the needed experience -- and contribute to the research process -- by serving on the advisory boards of academic departments, industry research labs, or specific projects or by otherwise interacting with the research community. Some leading companies are doing this already and are venturing into research in their own IT departments. The insight end users can provide into the needs of particular applications and their experience with operational systems will prove invaluable to IT research. Their participation will become increasingly necessary to assuring that their needs are met.

Finally, companies with established research and development divisions must attempt to engage end users more actively in IT research. Companies like IBM and Microsoft have already established novel programs for doing so, but many others need to take similar steps. These companies have a tradition of research and, in this respect, are well-positioned to have an immediate impact. Advancing understanding of large-scale systems and social applications should ultimately contribute to their bottom lines.

The committee believes that these recommendations, if implemented, will set the nation on a course that will allow it to both create the information technologies of tomorrow and make the most appropriate and beneficial use of them.

My colleagues and I will now take your questions. Since this briefing is being recorded, please tell us your name and affiliation when asking a question. Thank you.

(end transcript)

This site is produced and maintained by the U.S. Department of State. Links to other Internet sites should not be construed as an endorsement of the views contained therein.