|
PITAC - Report to the President
Creating an Effective Management Structure for Federal IT R&D |
 
|
|
|
|
|||
 |
|
||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
5. Creating an Effective Management Structure for Federal Information Technology R&D
|
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
5. Creating an Effective Management Structure
|
Creating an Effective Management Structure for Federal Information Technology R&D Given the importance of information technology to the Nation's future, it is critical that the Federal research investment in this area be well managed. An effective management structure for information technology research can ensure that the Federal research portfolio is appropriately balanced, can encourage innovative long-term research, can help the Administration articulate research priorities, and can help assure Congress that its goals are being met. The management structure should ensure a diversity of funding mechanisms, including single-investigator grants, multiple-investigator projects within and across disciplines, and integrated research centers. Such diversity in approach and funding modes will increase opportunities for unexpected discoveries, encourage a broader attack on problems, and ensure fewer missed opportunities. Unlike other disciplines that are usually incremental and evolutionary, computer science and engineering is a transformational and revolutionary technology. Since tomorrow's applications will be ones that we cannot even envision today, it is critical that we as a Nation invest in core computer science and engineering to enable those revolutionary advances. The Committee feels strongly that information technology research must not be driven entirely according to the extrapolated demands of today's commercial, scientific, and/or national security applications. An effective management structure should accomplish the following:
|
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
|
5.1 Agency Roles
|
Agency Roles Currently, funding for information technology R&D comes from several different agencies, with no single agency having information technology as its primary responsibility. This system has worked surprisingly well in the past and the Committee views the diversity of research funding as a crucial strength. However, the Federal funding portfolio has not kept pace with the importance of information technology to the Nation, especially with respect to support for fundamental research in computer science and engineering. The Committee recommends that the proposed budget increase be used primarily to fund fundamental information technology R&D, with a significant portion of the funds directed to the National Science Foundation (NSF). The Committee believes that NSF should use this increased funding to reinvent and significantly enhance its traditional role as the champion of core computer science and engineering research. In parallel, other agencies conducting information technology research should renew their commitment to fundamental research and engineering, with specific emphasis on making software a substantive component of every major information technology initiative. For example, the Defense Advanced Research Projects Agency's (DARPA) flexible and creative management style and its technical expertise and history of successful investments in information technology R&D has and should continue to make significant contributions. Participation by a variety of agencies, which tend to have different styles and emphases for supporting research or procuring technology, is important to establish and maintain an effective U.S. information technology R&D program. This diversity allows many views to compete and fosters pursuit of intellectual questions via a range of complementary modes. Such information technology research will be motivated by long-term mission goals, but should be evaluated and funded with fundamental information technology challenges in mind -- developments that will generalize to other domains despite their genesis in a specific mission agency. The Committee believes, for example, that biomedically motivated basic research in information technology should be supported by the National Institutes of Health (NIH) and viewed both as important information technology research and as fundamental biomedical research. Similar arguments can be made for DARPA, DOE, NASA, and the other research agencies that will be continue to be involved with mission-motivated basic research in information technology. Recommendation: Strongly encourage NSF to assume a leadership role in basic information technology research. Provide NSF the necessary resources to play this role. In its leadership role, NSF should foster interagency collaborations to ensure adequate funding levels in basic research. While NSF can only control its own budget, it should work with other agencies with similar research needs to assess major requirements and needs being addressed, identify shortcomings of the fundamental research portfolio, and initiate efforts to address deficiencies. This leadership role for basic information technology research may require innovations internal to the NSF, as may be necessary to ensure that NSF is able to define, support, and coordinate a broad range of modes of research support. Example modes of support are centers of diverse sizes and multiple-investigator projects with longer terms. NSF may also have to make changes to ensure that it fosters high-risk, high-payoff long-term research and invests sufficiently in the core of computer and information science and engineering. We also believe that there needs to be significantly more information technology representation on the National Science Board than exists now. To successfully carry out the program of research proposed in this document, NSF will need to support a portfolio of modes of research that differ from the current mix of mostly single-investigator research grants together with a small number of centers. For example, multi-investigator projects of longer duration (5-7 years of funding) are needed in order to carry out high-impact experimental research agendas, whether or not such grants are available in other parts of the Foundation. Some of these projects will include long-term basic technology research that has an engineering orientation and involves explicit technology transfer. Some projects should be supported in response to topic-specific solicitations in order to support a DARPA-like concentration of effort on compelling research areas. Because of the nature of the field, the portfolio must successfully meld science investigations with technology exploration. Roughly 40-50% of the proposed budget increases for information technology should go to NSF, to support basic research in information technology focused in the Computing and Information Science and Engineering (CISE) Directorate, with the rest allocated to other research agencies. The majority of the NSF increase should go to the new programs and modes of funding; the remainder should go to the traditional programs within the CISE Directorate, expanded as appropriate to projects of larger size and longer duration. |
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
|
5.2 Policy and Coordination of Information Technology R&D
|
Policy and Coordination of Information Technology R&D Given the importance of information technology to Nation's economy, security, and well being, the Committee believes there should be high-level management attention and focus by the Administration for information technology R&D. Recommendation: Designate a Senior Policy Official for Information Technology R&D. Information technology is of critical importance to the Nation; it is the basis for a $700B domestic industry that drives many of today's innovations and scientific discoveries and offers even greater potential for innovation in the next century. The Committee believes that information technology R&D requires a high-level policy voice, similar to that afforded to other priority research areas in the White House. Positioning an individual at this level would help ensure the Administration receives timely advice on information technology issues and foster a strategic approach in determining R&D investments. This individual would be responsible for leading the White House effort to establish Federal policies to support, encourage, and help coordinate long-range information technology development to maintain U.S. leadership in this vital part of our economy. Recommendation: Establish a senior-level policy and coordination committee to provide strategic planning and management. The principal goal of coordination is to ensure that the Federal information technology research portfolio is sufficiently aggressive and properly balanced to meet the pressing needs of the Nation. Since many Federal agencies fund information technology R&D, a high-level venue for cross-agency coordination is desirable. The coordinating committee should consist of agency directors, since it is they who have policy and budget making authority. This coordination committee should advise and report directly to the President's Assistant for Science and Technology to increase the overall effectiveness and productivity of Federal R&D efforts to develop and apply information and communications technology. The coordination committee should address significant national policy matters that cut across agency boundaries. The committee should establish objectives for research programs and review them to ensure that they are meeting those objectives. In addition, it should ensure that the overall Federal program is well-balanced and has good coverage of important topics. This coordination should include the entire Federal information technology R&D endeavor, including the efforts resulting from the proposed budget increase. A problem with the present management structure is the difficulty of getting a clear and complete picture of the total Federal investment in information technology, since there are Federal computing, information, and communications R&D programs that are not part of the formally coordinated Computing, Information, and Communications (CIC) programs. The coordination in information technology should encompass all major investments in R&D, as well as investments in people, supplies, and the like. For instance, high-end computing acquisitions are a necessary part of a coordinated program. Recommendation: Extend the HPCC program coordination model to major Federal information technology R&D activities. To make this coordination successful, the entire enterprise should have strong staff support in the White House and in the agencies themselves. The HPCC program, with the National Coordination Office and working groups with agency representation, is an effective model of interagency collaboration. The Committee recommends extending this model to all major Federal information technology R&D activities, with the NCO facilitating interagency coordination and supporting the management structure recommended by the Committee. |
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
|
5.3 Support and Implementation
|
Support and Implementation To achieve the goal of reinvigorating long-term, high-risk research in information technology, Federal funding agencies will need to support many different kinds of projects, including single-investigator grants, multiple-investigator projects within and across disciplines, and integrated research centers. To encourage the visionary planning needed to produce true innovation, funding for these activities will need to be provided for longer periods -- long enough for high-risk projects to be carried through to completion. Furthermore, because effective research in computer science involves collaborations of all sorts -- with other computer scientists, with researchers in other disciplines and at other institutions, with industry, and with the community -- the modes of support should encourage these collaborations, without turning them into strict requirements. Research projects should have the flexibility to organize themselves in the best ways to achieve the goal of technological innovation. Recommendation: Diversify the modes of research support to include more projects of broader scope and longer duration, placing a renewed emphasis on research carried out in teams. During the 1970s and 1980s, DARPA computer science researchers were encouraged to imagine dramatically different futures and to carry out projects that would explore those futures. Researchers were given enough resources and time so they could concentrate on problems rather than worry about their next proposals. The results were dramatic advances in speech recognition, robotics, chip design, high-performance computing, machine vision, artificial intelligence, and virtual reality. The Committee would like to see this spirit renewed and replicated across all Federal information technology R&D programs. Currently, most NSF fundamental research support is for single-investigator grants. Although this kind of support will remain an essential component of the research portfolio, information technology can often make the most dramatic progress when carried out in teams of moderate size. DARPA has used this model effectively for several decades, and it is rapidly becoming part of the portfolios of other funding agencies including NSF. The committee believes that team research in information technology should be at least as important as single-investigator research, including the National Science Foundation. Furthermore, inter-institutional team projects help build a coherent national research community. Recommendation: Fund collaborations with applications to drive information technology research, but take measures to ensure that research remains a primary goal. A principal goal of information technology research is to produce results that will someday have commercial or social applicability. Thus, it is important to understand applications to which information technology may be applied and to use this knowledge to drive research into new technological strategies. Therefore, the Committee recommends that the Federal Government continue to fund collaborations between information technology researchers and applications developers. However, the funding programs must be carefully structured to ensure that the research remains a principal focus. In the early 1990s, a principal focus of the Federal High Performance Computing and Communications (HPCC) Program was the use of large-scale applications efforts to drive research on parallel computing. In this vision, teams of applications developers and computer scientists would have worked together to rebuild major applications on parallel computing platforms developing many useful computational support technologies. Unfortunately, many of the projects came under pressure to meet progress and performance deadlines for the applications, leading them to jettison all but the most immediately applicable computer science research. To reemphasize long-term fundamental research, we must acknowledge that the time to applicability will be too long in many cases to directly affect an application project that is driving the research. For example, if a computer architecture researcher works with a scientist on computational fluid dynamics to understand how to design better computers for that application, it is unlikely that machines of the resulting architecture will run that specific application in a two to three-year time frame. In a funding program designed to foster long-term research, this situation must be completely acceptable. Indeed, it is essential that the funding programs be designed to give equal value to the information technology research as to progress on the application. Recommendation: Fund centers for Expeditions into the 21st Century. Projects of moderate size cannot alone address the entire spectrum of long-term research needed to sustain our national leadership. To foster research with truly dramatic impacts, the Committee recommends the creation of two types of center-sized activities: "Expeditions into the 21st Century," which will endeavor to intersect with the long-term technological future, and "Enabling Technology centers," which will focus on research with application to problems of national importance. "Expeditions into the 21st Century" will be centers, perhaps virtual, that bring together scientists, engineers, and computer scientists from academia, government, and industry to "live in the technological future." The mission of these expeditions will be to report back to the Nation what could be accomplished by using technologies that are quantitatively and qualitatively more powerful than those available today. In essence, these centers will create "time machines" to enable the early exploration of technologies that would otherwise be beyond reach for many years. Just as the Lewis and Clark expedition opened up our Nation and led to unanticipated expansion and economic growth, the ideas pursued by information technology expeditions could lead to unexpected results and nourish the industry of the future, creating jobs and benefits for the entire Nation. There are existing successful examples of the use of this "living in the future" approach. In the private sector one of the most famous examples is the Xerox Palo Alto Research center (Xerox PARC) where researchers created an experimental network of computers for use by individuals. This effort pioneered many of the revolutionary technologies that led to today's personal computers, including graphical user interfaces, pointing devices, laser printing, distributed file systems, and WYSIWYG word processing. In the university community, the MIT Media Lab has been conducting similar explorations. Finally, there is the example of ARPAnet, which evolved into today's Internet. The Committee recommends funding several Expeditions, each with a different focus. The focus may be on either a discipline-based theme, such as bioinformatics or multi-scale engineering, or on an infrastructure-based theme, such as distributed databases or tele-immersion. To establish a context, each Expedition should be based on assumptions not true today, for example, ubiquitous computing or a vast amount of simulation, as described in Gelernter's Mirror Worlds. Each Expedition need not be limited to a single such assumption, but an Expedition should invest sufficient resources to make exploration of its assumption areas, those parts of the map of the future, possible. Each Expedition would be required to carry out several activities, including:
Recommendation: Establish a program of Enabling Technology centers. The Committee also recommends establishment of centers of excellence in computer science and engineering research applied to particular applications of information and communications technology. These Enabling Technology centers (ETCs), located at university and/or Federal research institutes, will provide integrated environments for academia, industry, and Government to focus on the application of next-generation information technology to important national problems. There are many application domains where information and communications technologies could make a difference. These include computational science and engineering; health care; delivery of Government services/Digital Government; crisis management; environmental monitoring; life-long learning; law enforcement and public safety; arts, culture, and the humanities; intelligent transportation systems; improving the quality of life for persons with disabilities; and distributed work (e.g. telecommuting, collaboration by geographically distributed teams). Enabling Technology centers should be focused on applied technology and development. Researchers at the centers would conduct R&D on information technology problems arising from the center's application domain, develop new curricula for students and mid-curricula for both students mid-career professionals, participate in testbeds, and identify barriers to more widespread adoption of information technology in the application domain of concern. In addition to research and development, these centers should perform several functions, including:
Grants need to be of sufficient duration and size to support a critical mass of researchers interested in a particular applications domain. The Committee recommends each center be funded on the Science and Technology center funding model. The full term of an ETC would be 10 years. Annual funding of up to $10M per center is recommended, with up to 15 centers simultaneously in operation. Competitions should be held every three years. |
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
|
5.4 Annual Review
|
Annual Review Recommendation: Establish an annual review of research objectives and funding modes. Both the Coordination Committee and the President's Information Technology Advisory Committee (PITAC) should be instrumental in reviewing research objectives. PITAC's role in advising the President through NSTC serves to provide high-level private sector advice on information technology. The Coordination Committee will provide high-level advice from within the government. The Coordination Committee, with advice from the Advisory Committee, should conduct an annual review of research programs to ensure that they are achieving the goals set out for them. In particular, the Coordination Committee should ensure that the modes of support proposed for the programs -- centers, multiple-investigator interdisciplinary projects, and testbeds, along with a renewed emphasis on software and long-term research -- are not being compromised and that these modes of research support are meeting the goals set out for them. This review is intended to serve as a high-level check to ensure the Federal R&D portfolio is properly balanced, comprehensive, and properly coordinated. To maximize the opportunities for full and frank exchanges among the principals, the degree of formality and process associated with the preparation for and conduct of this review should be strictly limited. Given that the purpose is to coordinate a scientific endeavor, the review might more closely resemble a scientific workshop than a traditional committee meeting. |
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||
|
5.5 Funding
|
Funding Budget Recommendation: Increase current funding for IT R&D as follows over the fiscal years 2000-2004. Table 5.1 summarizes the recommended funding increments for each of the four research priority areas identified in this report. Descriptions of the research topics and associated funding tables were provided in Chapters 3 and 4 of this report. The funding recommendations are in addition to those for current (FY 1999) Federal IT R&D programs and represent, rather, new or expanded activities. Table 5.1 Recommending funding increases for IT R&D ($ in millions)
|
||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||
|
Can the Proposed Increases be Invested Effectively? One question that arose during the public comment on the interim report was whether the IT research community could effectively invest an increase of the size we are proposing. To answer that question, we have developed a simple investment model described here. Note that this is not a proposal on how invest the funds, but rather an exercise to determine whether we can reasonably expect the community to be able to take on the challenge. There are three principal ways that funding increases might be used in to affect change in the research community:
Our analysis also establishes that there is unused research capacity in the existing research community. Currently the top-rated universities have support levels per researcher in the range of $400K to $500K per year, while universities not ranked in the top five have support levels substantively lower, in the range of $150K per researcher per year. One goal might be to bring the next twenty-five departments up to the quality and support levels of the top five and another fifty departments up to about $250K per researcher per year. Assume that 25 organizations (with about 40 principal researchers per organization) will need to have their support base increased from about $150K to about $500K per principal researcher. We also assume that the next 50 organizations (also with about 40 principal researchers per organization) will need to have their support base increased from about $100K to $250K per principal researcher. These target levels are quite reasonable in terms of today's costs. For example, $250K will typically cover the cost of two graduate students, a programmer, some equipment, and summer salary for the researcher. The above calculation leads us to estimate an expenditure of $650M per year after five years. These numbers reflect the vital importance of upgrading the current research base to improve the overall effectiveness of current funding and to insure that the universities remain competitive with the IT industry in providing computational and networking environments sufficient to be attractive to new graduate students. Without this latter investment, the needed increase in researchers may not materialize in the universities. These investments will insure that the universities can attract the needed researchers and that the existing research base can be made competitive with the state-of-the-art in industry. Following the recommendations of the report, we have budgeted $330M for large shared infrastructure such as supercomputers, external networks and testbeds to link the researchers and associated maintenance. This leaves us with the following totals, which are consistent with the budgets proposed in the report, for the fifth year:
|
|||||||||||||||||||||||||||||||||||||||||||
|
|
||||||||||||||||||||||||||||||||||||||||||
|
|
|
||||||||||||||||||||||||||||||||||||||||||
|
|
|||||||||||||||||||||||||||||||||||||||||||