Subcommittee on Research of the House Committee on Science

Testimony of Cita M. Furlani
Director, National Coordination Office for IT R&D

• Immediate on-site medical care, in the home and at remote locations
• Reliable, failure-resistant systems for such mission-critical applications as air-traffic control, financial transactions, life support, and power supply
• Industrial process and product modeling, visualization, and analytical capabilities, such as in aircraft design and production, automotive efficiency and safety, and molecular synthesis of new drugs
• Expanded e-commerce with assured security and privacy of information
• On-demand universal access to education and knowledge resources
• Advanced computing capabilities that underpin the Nation's leadership in science and technology, including the biotechnology revolution, and the success of critical civilian and national security missions of the Federal government
• More accurate weather forecasting and improved environmental analysis and decision making
• High-performance networking and information systems for emergency and disaster management
• Access to information anytime, anywhere, with any device

Goals of the Federal investment
Federal IT research helped fuel the computing revolution and the unprecedented U.S. economic prosperity of recent years. And it continues to spur major technological innovations in computing and networking, such as the development of optical networking and experimentation with nano-scale and quantum computing and conducting materials. These are results of the Federal effort to accelerate development of the underlying technologies - called "enabling" technologies - on which all computing and networking devices and systems are based. By accelerating advances in these fundamental, underlying technologies, IT R&D enables Federal agencies to accelerate their development and deployment of state-of-the-art advanced systems and applications needed for critical government missions. Federal IT R&D also is devoted to readying laboratory advances for deployment and strengthening experimental technologies through prototyping, testbeds, and evaluations.
 
The multiagency program
I commend the Congress for its farsightedness in enacting the High Performance Computing (HPC) Act of 1991 (P.L. 102-194), which mandated a multiagency research effort to carry on the unique Federal role in long-range IT R&D. Not only is this program scientifically vital, but it offers a remarkably successful example of effective collaboration among Federal agencies and with the private sector. The bipartisan HPC legislation established a powerful framework for Federal IT research, combining ambitious research goals with specific requirements for interagency cooperation, coordination, and partnerships with academe and industry. That framework has evolved over a10-year period into a very productive research enterprise that involves all the major Federal science and research agencies and collaborations with virtually all major U.S. research universities and with many companies involved in developing new information technologies and applications. Technical and administrative support for the program is provided by the National Coordination Office for Information Technology Research and Development, which I direct.
 
Brief history of the program (see figure 1)
The10-year history of the Federal multiagency IT research effort is partly one of changing nomenclature and partly one of burgeoning research programmatic interests in this explosively growing and increasingly critical scientific field. These interests expanded the program's focus from high performance computing and networking to applications that make advanced computing and networking capabilities more widely available and easier to use. Beneath the name changes and the programmatic evolution, however, the structure and budget process for the program have not changed. Most significant, the focus of the Federal effort has remained constant to the Congress's wise intent: To conduct the long-term, fundamental research that leads to technological breakthroughs advancing the science of information technology. That is the kind of work that made the U.S. the world leader in advanced computing and networking in the first place. And it is the kind of work required to sustain that preeminence over the long term.
 
From FY 1992 to FY 1996, this program was called the High Performance Computing and Communications (HPCC) Program. In FY 1997, it was renamed the Computing, Information, and Communications (CIC) programs. The programs' research groups, which had been called Components, were renamed Program Component Areas (PCAs) and were updated to reflect the growth of research needs beyond the HPCC Program's initial focus on high-end computing and advanced networking. In FY 1998, the Congress enacted a three-year Next Generation Internet (NGI) Initiative, which was included as part of the CIC budget crosscut. In FY 2000, the Administration sponsored an "Information Technology for the 21st Century (IT2) Initiative. In FY 2001, both the NGI and IT2 initiatives were incorporated into the renamed Information Technology R&D (IT R&D) Program. In the President's FY 2002 Budget, the program is referred to as the Networking and Information Technology Research and Development (NITRD) program.
 
Participating agencies
The participating agencies are, in alphabetical order:

Agency for Healthcare Research and Quality (AHRQ)
   of the Department of Health and Human Services
Defense Advanced Research Projects Agency (DARPA)
   of the Department of Defense (DoD)
Department of Energy (DOE) National Nuclear Security
   Administration (NNSA)
DOE Office of Science
Environmental Protection Agency (EPA)
National Aeronautics and Space Administration (NASA)
National Institutes of Health (NIH) of the Department
   of Health and Human Services
National Institute of Standards and Technology (NIST)
   of the Department of Commerce
National Oceanic and Atmospheric Administration (NOAA)
   of DOC
National Security Agency (NSA) of DoD
National Science Foundation (NSF), and
Office of the Deputy Under Secretary of Defense
   for Science and Technology (ODUSD [S&T]).

The Congress appropriates funding for the multiagency IT research program through agency appropriations, and this funding is reported in a budget crosscut. References in this testimony to "participating agencies" refer to agencies included in the budget crosscut. Other agencies, such as the Federal Aviation Administration and the General Services Administration, participate in NITRD discussions but are not part of the budget crosscut.
 
Program structure (see figure 2 on next page) The Congress's prescient requirement for collaboration in the HPC Act is particularly significant because information technology itself is an intensive interdisciplinary scientific endeavor in which collaboration across many science and engineering fields is a necessity. The alignment of scientific and programmatic imperatives for cooperative work in the Federal IT research effort has stimulated precisely the kind of ongoing exchange of ideas and cross-agency initiatives that the Congress envisioned.
 
Let me describe how this works logistically in the research effort that has evolved out of the High Performance Computing and Communications (HPCC) Program established by the Congress in 1991.
 
First, the White House sets broad policies and priorities for Federal networking and information technology R&D - from the President through the Office of Science and Technology Policy (OSTP) and the National Science and Technology Council (NSTC). In addition, the President's Information Technology Advisory Committee (PITAC), the group of IT leaders in industry and academe that provides independent advice to the President on IT issues, periodically makes recommendations about Federal R&D that influence NITRD directions.
 
The hands-on coordination of interagency IT research activities is handled by the Interagency Working Group (IWG) on IT R&D. This group is made up of representatives from each of the participating agencies and from the Office of Management and Budget (OMB), OSTP, the National Economic Council (NEC), and my office, the National Coordination Office for IT R&D. The IWG is chaired by Dr. Ruzena Bajcsy, Assistant Director of the National Science Foundation and head of NSF's Directorate for Computer and Information Science and Engineering.
 
The major research emphases of the NITRD effort are reflected in the Program Component Areas (PCAs). The work of each PCA is guided by a Coordinating Group of program managers from participating agencies. These groups meet monthly to coordinate the objectives and activities of the multiagency projects in their specialized research areas. They report to the Interagency Working Group (IWG). The PCAs evolve in response to changing research needs.

The PCAs are:

• High End Computing (HEC), which includes both HEC R&D and HEC Infrastructure & Applications (I&A Human Computer Interaction & Information Management (HCI&IM)
• Large Scale Networking (LSN)
  

  LSN also has three subordinate teams - High Performance Networking Applications Team (HPNAT), Joint Engineering Team (JET), and Networking Research Team (NRT) - that address specific technical issue areas. These teams include non-Federal members from academe and industry.

• Software Design and Productivity (SDP)
• High Confidence Software and Systems (HCSS)
• Social, Economic, and Workforce Implications of IT and IT Workforce Development (SEW)

• The annual Supplement to the President's Budget, known as the Blue Book, documents the current year's research accomplishments and proposed scope of work for the next fiscal year, with budget estimates for the current year and budget requests for the next fiscal year, by agency and by PCA.
   
  
• A more detailed mapping of PCA activities to agencies' internal programs and budgets is prepared in the annual Implementation Plan for the multiagency budget crosscut.
   
  
• Research planning papers periodically prepared by the Coordinating Groups provide a basis for discussion of research focuses and approaches within each PCA and across the NITRD portfolio. Often, these documents are developed from national research workshops sponsored by the Coordinating Groups, providing a forum for discussions with experts from academe and industry about key issues and future goals in IT R&D. Reports of these meetings also are published and become part of the ongoing assessment of research directions.

• Federal IT R&D occupies a unique R&D niche. It is the Nation's primary source of long-term, fundamental research on IT issues that must be addressed to advance the capabilities of computers, networks, and information systems generally. Industry and venture capitalists typically focus on short-term product development offering the likelihood of rapid returns on investment. Private-sector investment strategies therefore tend to bypass key technology areas that may be the most critical to Federal government missions and that help support the continuing superiority of the U.S. IT industry. These areas include high-end computing, mass storage, optical networking, interoperable systems and applications, security, privacy, new generations of embedded and large scale systems, improved processes for developing new software, and effective human uses of IT. In fundamental IT R&D, the research time horizons are much longer and there is no guarantee of the success of any one research path. IT industry leaders have been for many years among the most ardent champions of Federal investment in long-term, fundamental IT research - precisely because the U.S. government is a primary source of support for that kind of risky, pre-competitive exploration. It is not necessarily the case that the private sector should not have a strong role in the support of fundamental research, but the difficulty to the company in appropriating the returns makes commercial support of this research more difficult.
   
  
• Federal IT R&D supports critical agency missions and national needs, including national defense and national security, critical infrastructure protection, energy systems, aerospace engineering, weather and climate forecasting, and advanced biomedical and other scientific research. National defense and national security needs alone require advanced IT research efforts on a continuing basis to equip the military with cutting-edge weapons technologies and secure communications systems and to accurately model and design these advanced systems.
   
  
• The research portfolio is diversified and balanced. Federal research responds to a basic reality of the interdisciplinary IT field: What can be accomplished using IT is determined by the weakest or the slowest technology, not by the strongest or the fastest. For that reason, Federal IT R&D pursues a balanced, diversified portfolio of research interests, seeking advances across the wide range of enabling technologies required for agency missions.
   
  
• Federal IT R&D produces broadly useful technologies and tools that spur innovation across the U.S. economy. It is an effective engine of technology transfer. The coordination of Federal IT R&D investments across many agencies and private-sector partnerships leverages mission-related research, producing general-purpose, broadly useful, and interoperable technologies, tools, and applications. Federal IT R&D has thus been a powerful engine of technology transfer, the direct result of its focus on widely applicable solutions to basic IT problems and its mechanisms of funding R&D. The large number of Federally funded breakthroughs subsequently commercialized in the private sector - often by graduates of U.S. research universities whose education was funded through the IT R&D programs - leverage the Federal investments even further.
   
  
• The Federal IT R&D program maintains the talent pipeline the Nation needs to continue making technological advances. More than half of the NITRD budget is expended for university-based research. This work is peer-reviewed, which helps assure it is of the highest quality, and the research dollars help support the graduate training of new generations of highly skilled scientists, engineers, and technicians needed to work on national research challenges.

• The multiagency approach aligns well with the multidisciplinary nature of IT research issues.
   
  
• The NITRD effort involves the right Federal agencies. The program also draws in a variety of other interested agencies to participate in its meetings, research workshops, and other activities. Such agencies include the Bureau of Labor Statistics, the Department of Defense High Performance Computing Modernization Office, the Federal Aviation Administration, and the General Services Administration.
   
  
• Through interactions in this program, the NITRD agencies have developed active, trusting, and efficient relationships for communication and cooperation. For example, agencies now participate in each other's proposal review processes.
   
  
• The missions of the participating agencies span all major IT R&D areas, ranging from fundamental research in broadly useful enabling technologies to applying those technologies to a wide variety of applications. NSF invests in fundamental research itself; some agencies invest in fundamental research in order to accomplish their agency's unique goals; and others use the technologies developed by these agencies to conduct R&D on their own agency's mission-driven applications. Individually, the agencies could not have accomplished all that the multiagency effort has made possible. A special by-product of this collaboration is that NITRD technologies are able to interoperate, which benefits both Federal operations and U.S. economic competitiveness. So, for example, today each agency does not have its own stand-alone network. Rather, the Federal government and the country have the Internet, a network of networks that is still evolving. Agencies are focusing on complementary efforts, making the coordinated program substantially greater than the sum of its agency parts.
   
  
• The Program Component Area system of organizing NITRD research emphases provides the program with the flexibility to evolve quickly in response to changing R&D needs. For example, even the PITAC's compelling 1999 recommendations for IT research made no mention of one key area - microsensors and embedded devices -that is now of paramount importance for national defense. The PITAC report made only passing reference to a second area - assurance and security in software - that is now viewed as a prerequisite for development of any mission-critical software. The PCAs are able to evolve their interests, as noted above, in a timely way to encompass such emerging research focuses.
   
  
• The multiagency coordination process has facilitated innovative collaborative research efforts that individual agencies would not otherwise have been able to tackle. A notable example is the Large Scale Networking program, in which NSF, DARPA, NASA, NIH, DOE National Nuclear Security Administration, DOE Office of Science, NIST, NOAA, AHRQ, and ODUSD (S&T) collaborate in networking research and development. Since1998, LSN research has boosted the end-to-end performance of shared end-user applications over networks from 1.5 megabits per second to 1 gigabit per second today. That is nearly a thousand-fold improvement in the end-to-end performance of networks for the Nation's scientific research community.
   
  
• Collaborative efforts include industry partnerships. Critical networking integration activities in the LSN effort, for example, involve such corporations as AT&T, Cisco, MCI Worldcom, Nortel, Qwest, and Sprint. Representatives of Internet2 participate in the LSN specialized teams. Intel, Sun, and others are developing an automatically tuned application/network interface; Ciena and others are developing key optical networking components. NSF partnered with Compaq in the research to develop the terascale computing platform at the Pittsburgh Supercomputing center. The terascale machine is the newest and most powerful addition to NSF's Partnerships for Advanced Computational Infrastructure (PACI) program. DOE/NNSA is collaborating with more than 20 organizations representing academia, industry, and government to develop a next-generation storage system based on commercially available products. Types of industry involvement include: CRADAs, collaboration, IPAs, consortia, start-ups, tech transfer, procurement, standards development, and advisory committees.
   
  
• The NITRD program is responsive to constituent research communities in universities and in industry. Its focused workshops, for example, draw the attention and participation of academic and industry researchers nationally and help shape or revise the research agenda in specific areas. Recent workshops on future directions in networking and on long-range issues in software design and productivity were well attended and enthusiastically endorsed by leading academic and industry experts in these fields.
   
  
• The Federal IT research investment has consistently received bipartisan congressional support and the support of the Executive Branch.

• Undertaking the R&D necessary to tackle IT problems at scale. For example, we have never built a network that is the size and complexity of the Internet. We do not know how to do that today. But when we think we do know how, we will need to experiment, test, and evaluate the model in order to make it work and work well. We will need large-scale R&D testbeds for this research.
   
  
• Providing infrastructure for research on high-end applications for the sciences and engineering (for example, growing the NSF Partnerships for Advanced Computing Infrastructure [PACI]). Access to high-performance computing platforms and high-speed networks remains very limited today, constraining the ability of researchers to build and test the most advanced scientific applications. Access to computing cycles at the NSF supercomputing centers has declined as the demand for computing time increases.
   
  
• Working with other Federal agencies on their IT problems. Some agencies do not have a research mission, but the IT transformation that they are undergoing requires that they take on new tasks, or tackle old tasks in new ways. This requires research, and the implementation of research-based innovations. The NITRD agencies are willing to work with other agencies on these issues.
   
  
• Helping the Congress and the U.S. public better understand how IT advances are contributing to advances in not just the sciences but also the humanities. Many benefits could flow from greater public understanding of IT, ranging from increased student interest in science, mathematics, and research, to broader awareness of the impacts of information technology on all of us.