<DOC>
[108th Congress House Hearings]
[From the U.S. Government Printing Office via GPO Access]
[DOCID: f:97999.wais]
DEFINING FEDERAL INFORMATION TECHNOLOGY RESEARCH AND DEVELOPMENT: WHO?
WHERE? WHAT? WHY? AND HOW MUCH?
=======================================================================
HEARING
before the
SUBCOMMITTEE ON TECHNOLOGY, INFORMATION
POLICY, INTERGOVERNMENTAL RELATIONS AND
THE CENSUS
of the
COMMITTEE ON
GOVERNMENT REFORM
HOUSE OF REPRESENTATIVES
ONE HUNDRED EIGHTH CONGRESS
SECOND SESSION
__________
JULY 7, 2004
__________
Serial No. 108-251
__________
Printed for the use of the Committee on Government Reform
Available via the World Wide Web: http://www.gpo.gov/congress/house
http://www.house.gov/reform
U.S. GOVERNMENT PRINTING OFFICE
97-999 WASHINGTON : 2004
_________________________________________________________________
For sale by the Superintendent of Documents, U.S. Government Printing
Office Internet: bookstore.gpo.gov Phone: toll free (866)512-1800;
DC area (202) 512-1800 Fax: (202) 512-2250 Mail: Stop SSOP,
Washington, DC 20402-0001
COMMITTEE ON GOVERNMENT REFORM
TOM DAVIS, Virginia, Chairman
DAN BURTON, Indiana HENRY A. WAXMAN, California
CHRISTOPHER SHAYS, Connecticut TOM LANTOS, California
ILEANA ROS-LEHTINEN, Florida MAJOR R. OWENS, New York
JOHN M. McHUGH, New York EDOLPHUS TOWNS, New York
JOHN L. MICA, Florida PAUL E. KANJORSKI, Pennsylvania
MARK E. SOUDER, Indiana CAROLYN B. MALONEY, New York
STEVEN C. LaTOURETTE, Ohio ELIJAH E. CUMMINGS, Maryland
DOUG OSE, California DENNIS J. KUCINICH, Ohio
RON LEWIS, Kentucky DANNY K. DAVIS, Illinois
JO ANN DAVIS, Virginia JOHN F. TIERNEY, Massachusetts
TODD RUSSELL PLATTS, Pennsylvania WM. LACY CLAY, Missouri
CHRIS CANNON, Utah DIANE E. WATSON, California
ADAM H. PUTNAM, Florida STEPHEN F. LYNCH, Massachusetts
EDWARD L. SCHROCK, Virginia CHRIS VAN HOLLEN, Maryland
JOHN J. DUNCAN, Jr., Tennessee LINDA T. SANCHEZ, California
NATHAN DEAL, Georgia C.A. ``DUTCH'' RUPPERSBERGER,
CANDICE S. MILLER, Michigan Maryland
TIM MURPHY, Pennsylvania ELEANOR HOLMES NORTON, District of
MICHAEL R. TURNER, Ohio Columbia
JOHN R. CARTER, Texas JIM COOPER, Tennessee
MARSHA BLACKBURN, Tennessee BETTY McCOLLUM, Minnesota
PATRICK J. TIBERI, Ohio ------
KATHERINE HARRIS, Florida BERNARD SANDERS, Vermont
(Independent)
Melissa Wojciak, Staff Director
David Marin, Deputy Staff Director/Communications Director
Rob Borden, Parliamentarian
Teresa Austin, Chief Clerk
Phil Barnett, Minority Chief of Staff/Chief Counsel
Subcommittee on Technology, Information Policy, Intergovernmental
Relations and the Census
ADAM H. PUTNAM, Florida, Chairman
CANDICE S. MILLER, Michigan WM. LACY CLAY, Missouri
DOUG OSE, California STEPHEN F. LYNCH, Massachusetts
TIM MURPHY, Pennsylvania ------ ------
MICHAEL R. TURNER, Ohio
Ex Officio
TOM DAVIS, Virginia HENRY A. WAXMAN, California
Bob Dix, Staff Director
Ursula Wojciechowski, Professional Staff Member
Juliana French, Clerk
Adam Bordes, Minority Professional Staff Member
C O N T E N T S
----------
Page
Hearing held on July 7, 2004..................................... 1
Statement of:
Fossum, Donna, manager, RaDiUS project, RAND Corp.; Edward
Lazowska, co-Chair, President's Information Technology
Advisory Committee and Chair, Department of Computer
Science and Engineering, University of Washington; William
Scherlis, professor, School of Computer Science at Carnegie
Mellon; and Stephen Squires, chief science officer, vice
president, Hewlett-Packard................................. 55
Nelson, Dr. David, Director, National Coordination Office for
Information Technology Research and Development (Executive
Office of the President); Dr. Peter Freeman, co-Chair of
Interagency Working Group and Assistant Director, Computer
and Information Science and Engineering Directorate,
National Science Foundation; Dr. Hratch Semerjian, Acting
Director, National Institute of Standards and Technology;
and Dr. C. Edward Oliver, Associate Director, Office of
Advanced Scientific Computing Research, U.S. Department of
Energy..................................................... 10
Letters, statements, etc., submitted for the record by:
Clay, Hon. Wm. Lacy, a Representative in Congress from the
State of Missouri, prepared statement of................... 8
Fossum, Donna, manager, RaDiUS project, RAND Corp., prepared
statement of............................................... 58
Freeman, Dr. Peter, co-Chair of Interagency Working Group and
Assistant Director, Computer and Information Science and
Engineering Directorate, National Science Foundation,
prepared statement of...................................... 24
Lazowska, Edward, co-Chair, President's Information
Technology Advisory Committee and Chair, Department of
Computer Science and Engineering, University of Washington,
prepared statement of...................................... 73
Nelson, Dr. David, Director, National Coordination Office for
Information Technology Research and Development (Executive
Office of the President), prepared statement of............ 14
Oliver, Dr. C. Edward, Associate Director, Office of Advanced
Scientific Computing Research, U.S. Department of Energy,
prepared statement of...................................... 41
Putnam, Hon. Adam H., a Representative in Congress from the
State of Florida, prepared statement of.................... 4
Scherlis, William, professor, School of Computer Science at
Carnegie Mellon, prepared statement of..................... 85
Semerjian, Dr. Hratch, Acting Director, National Institute of
Standards and Technology, prepared statement of............ 33
Squires, Stephen, chief science officer, vice president,
Hewlett-Packard, prepared statement of..................... 97
DEFINING FEDERAL INFORMATION TECHNOLOGY RESEARCH AND DEVELOPMENT: WHO?
WHERE? WHAT? WHY? AND HOW MUCH?
----------
WEDNESDAY, JULY 7, 2004
House of Representatives,
Subcommittee on Technology, Information Policy,
Intergovernmental Relations and the Census,
Committee on Government Reform,
Washington, DC.
The subcommittee met, pursuant to notice, at 10:30 a.m., in
room 2154, Rayburn House Office Building, Hon. Adam Putnam
(chairman of the subcommittee) presiding.
Present: Representatives Putnam and Clay.
Staff present: Bob Dix, staff director; John Hambel, senior
counsel; Ursula Wojciechowski, professional staff; Juliana
French, clerk; Felipe Colon, fellow; Michelle Ash, minority
senior legislative counsel; Adam Bordes, minority professional
staff member; and Cecelia Morton, minority office manager.
Mr. Putnam. A quorum being present, this hearing of the
Subcommittee on Technology, Information Policy,
Intergovernmental Relations and the Census will come to order.
Good afternoon and welcome to the subcommittee hearing on
``Defining Federal Information Technology Research and
Development. Who? Where? What? Why? and How much?'' the purpose
of this hearing is to examine the extent of Federal funding for
and the leveraging of information technology research and
development across agencies, academia and industry.
By addressing the basic questions, this subcommittee hopes
to identify the following: How many different agencies of the
Federal Government are currently engaged in conducting or
managing IT research and development; is there an overall
strategic plan that provides an opportunity to leverage
investments, both internally and externally, and to identify
complementary activities in an effort to avoid duplication; how
much is being spent on an annualized basis on information
technology R & D; where and how these investments are actually
being made; what are the outcome measurements and expectations
associated with those investments; is there a defined set of
goals and objectives or focus areas that are targeted by these
efforts and what have been the recent results; what is the role
of the academic community and the private sector, and how are
these partnerships created and maintained?
The Federal Government funds research and development to
meet the mission requirements of the departments and agencies.
Advances in the uses of IT research and development are
continuing to change the way those Federal agencies
communicate, use information, deliver services and conduct
business. The technology and expertise generated by this
endeavor may have applications beyond the immediate goals or
intent of federally funded research and development. Federal
support reflects the consensus that while basic research is the
foundation for many innovations, the rate of return to society
as a whole generated by investments and such work is
significant.
The potential benefits of federally funded R&D related to
information technology are endless. Federally funded programs
have played a crucial role in supporting long term research
into the fundamental aspects of computing. The unanticipated
results of research are often as important as anticipated
results. The Internet, electronic mail and instant messaging
were by-products of government funded research from the 1960's.
Another aspect of government funded IT R&D is that it often
leads to open standards, something that many perceive as
beneficial, encouraging deployment and further investment.
Previous oversight hearings conducted by this subcommittee
have identified an important missing link in the cyber security
arena that requires further attention in the research and
development area. We have learned that inadequate tools exist
today to conduct necessary quality assurance testing of
existing and emerging software and hardware products that could
better identify flaws, defects and other vulnerabilities prior
to deployment. With a renewed commitment on the part of
software and hardware manufacturers to quality and security of
the products they introduce into the marketplace, a
collaborative approach to developing more mature testing tools
are essential to improved protection of computer networks and
the information assets they contain.
The outcomes achieved through public and private funding
programs create a synergistic environment in which both
fundamental and application driven research is conducted,
benefiting government, industry, academia and the public.
Government funding appears to have allowed research on a larger
scale and with greater diversity, vision and flexibility than
would have been possible without government involvement.
It is important to recognize collaborative efforts across
programs and agencies and stress the importance of leveraging
efforts with academia and the private sector. Universities,
private companies, and Federal labs are important partners in
this endeavor. It will be productive to explore new methods to
encourage increased activities by other parties in the
innovation process, particularly if the goal is to continue the
technological advancement which has been so instrumental to
this Nation's economic growth and high standard of living.
Because investments in science and technology have resulted
in unparalleled economic growth as well as the standard of
living and quality of life, we must emphasize the importance of
supporting the efforts of IT R&D. Advances have been possible
only with the support of the public and private investment in
R&D, according to the President's budgets. Yet challenges
continue. There are many R&D needs vying for a limited amount
of R&D dollars. Federal research and development program
managers face tough choices in deciding where the money should
go and how much is appropriate for information technology.
Further, it is important to ensure that Federal agencies
are not pursuing conflicting goals. It is essential that
agencies, universities and industry move toward a more
coordinated, unified approach. Multiple Federal agencies will
need to coordinate their efforts to ensure that new
understanding of information technology and network security is
generated and that this knowledge is transitioned into useful
products. Academia will have developed and expanded degree
programs to ensure that an adequate work force exists to put
new tools and techniques into practice. The private sector has
a critical role to play, as it will contain the developers and
suppliers as well as the major purchasers of new IT
technologies and services.
Government sponsorship of research, especially in
universities, helps develop the IT talent used by industry,
universities, and other pieces of the economy. When companies
create products using the ideas and work force that results
from federally sponsored research, they repay the Nation in
jobs, tax revenues, productivity increases and global
leadership.
We need a strong strategic plan to ensure that IT R&D is
being used to maximize improvement and mission goals and
performance. federally funded research and development are key
endeavors within the respective agencies and in cooperation
with universities in the private sector. It is essential to
meet vital Federal needs and sustain global leadership in
science and in the engineering of information technology.
I welcome today's distinguished panel of witnesses and look
forward to their testimony and the opportunity to explore these
matters in greater detail.
At this time I would like to recognize the distinguished
ranking member of the subcommittee, Mr. Clay, for his opening
statement. Mr. Clay.
[The prepared statement of Hon. Adam H. Putnam follows:]
[GRAPHIC] [TIFF OMITTED] 97999.001
[GRAPHIC] [TIFF OMITTED] 97999.002
[GRAPHIC] [TIFF OMITTED] 97999.003
Mr. Clay. I thank the chairman for holding today's hearing
on what is an important but often overlooked portion of our
government's research and development portfolio. The Federal
Government will spend approximately $60 billion on the many
different components of information technology during fiscal
year 2004. In contrast, the fiscal year 2004 budget only
allocates $2.2 billion for the Networking and Information
Technology Research and Development Program, a minimal amount
considering the role of higher performance computing and
technology in our mission to enhance government efficiency,
accessibility and security for all citizens.
Although funding for IT research and development has
increased fourfold since 1990, along with an increased
coordination throughout multiple agency participants for such
activity, there is a disconnect between the level of government
funding and its importance in the development of a strong IT
work force and premier academic institutions.
Furthermore, the government's role in IT research and
development fosters the creation of common criteria and open
standards that both government and private industry can utilize
for their benefit. When focused, the government's investments
in IT research often results in jobs, economic growth, and a
higher standard of living in both quantitative and qualitative
terms. Moreover, such resources permits our Nation to remain on
the cutting edge of technology in vital areas, including health
care, education, manufacturing, and the basic sciences.
This concludes my remarks, Mr. Chairman, and I ask that
they be included in the record.
[The prepared statement of Hon. Wm. Lacy Clay follows:]
[GRAPHIC] [TIFF OMITTED] 97999.004
[GRAPHIC] [TIFF OMITTED] 97999.005
Mr. Putnam. Without objection, they will be included in the
appropriate place in the record.
At this time we will move to the administration of the
oath. If our witnesses would please rise and raise your right
hands.
[Witnesses sworn.]
Mr. Putnam. Note for the record that all the witnesses
responded in the affirmative, and we will move to our first
panel's testimony. I would ask that all of our panelists adhere
to the 5-minute rule for your opening statements and we will
have successive rounds of questions from the panel to get to
all of your issues.
Our first witness is Dr. David Nelson. Dr. Nelson is the
Director of the National Coordination Office for IT Research
and Development and a member of the Senior Executive Service.
He is responsible for the coordination of planning, budget and
assessment activities for the Federal networking and
information breakthrough that advance the science of IT.
Dr. Nelson is cochair of the Interagency Working Group for
the NITRD program. Dr. Nelson joined the NCO from NASA, where
he was Deputy CIO with primary responsibility for information
technology security of all NASA systems and additional
responsibilities in scientific computing and enterprise
architecture. He previously served at the Department of Energy,
which he joined from Oak Ridge National Labs, where he was
research scientist working mainly in theoretical plasma physics
and its applications to fusion energy. He is the author of
numerous papers in theoretical plasma physics, computational
science and research policy. He has twice received the
President's Meritorious Rank Award for superior sustained
managerial performance.
Welcome to the subcommittee. You are recognized for 5
minutes, Dr. Nelson.
STATEMENTS OF DR. DAVID NELSON, DIRECTOR, NATIONAL COORDINATION
OFFICE FOR INFORMATION TECHNOLOGY RESEARCH AND DEVELOPMENT
(EXECUTIVE OFFICE OF THE PRESIDENT); DR. PETER FREEMAN, CO-
CHAIR OF INTERAGENCY WORKING GROUP AND ASSISTANT DIRECTOR,
COMPUTER AND INFORMATION SCIENCE AND ENGINEERING DIRECTORATE,
NATIONAL SCIENCE FOUNDATION; DR. HRATCH SEMERJIAN, ACTING
DIRECTOR, NATIONAL INSTITUTE OF STANDARDS AND TECHNOLOGY; AND
DR. C. EDWARD OLIVER, ASSOCIATE DIRECTOR, OFFICE OF ADVANCED
SCIENTIFIC COMPUTING RESEARCH, U.S. DEPARTMENT OF ENERGY
Dr. Nelson. Thank you, Mr. Chairman and members of the
subcommittee. I have submitted my written testimony to the
subcommittee and ask that it be entered into the record, and I
will limit my oral testimony to a brief summary of four points.
Let me start by saying that I agree with many of your
opening comments, both from the majority and the minority side,
with regard to the importance of information technology
research and development, and I think that agreement will be
shown through my oral testimony.
First, I would like to discuss the Networking and
Information Technology Research and Development Program. This
program derives from authorization in the High Performance
Computing Act of 1991.
For fiscal year 2005 the President's budget requests
slightly over $2 billion for the program in 13 participating
agencies. The program supports long-range research as well as
research infrastructure, such as research computer centers and
research networks. Performers include universities, Federal
research centers and laboratories, national laboratories and
federally funded research and development centers, private
companies, and nonprofit organizations. Research is funded by
the participating agencies through grants, cooperative
agreements, contracts, and other authorities. The agencies work
together under the program to identify research needs, plan
research programs, and review progress.
I brought along one copy of planning research needs. This
is in the high confidence software and systems research area
and was developed by one of the coordinating groups under the
program.
Agencies may coordinate their selection of research
performers through joint solicitations and coordinated proposal
reviews. The program interacts with stakeholders through
workshops and other meetings and disseminates research results
through publications, reports and presentations. Often
activities under the program are conducted jointly with other
Federal programs that benefit from information technology.
Historical accomplishments include the High Performance
Computing and Communications Initiative in the early 1990's
that helped create modern computational science, parallel
supercomputers, the modern Internet and Mosaic, the first
graphical Web browser. The Next Generation Internet Initiative
in the late 1990's helped to create the technology for today's
high bandwidth optical networks and demonstrated the basis for
today's high performance network computing.
The program receives advice and guidance from the
President's Information Technology Advisory Committee, which
was authorized in the High Performance Computing Act. Members
of the committee are drawn from the private sector, and I
believe Dr. Ed Lazowska, cochair of the committee, is
testifying before the subcommittee today.
Let me turn to my second main point. This concerns the
value of the government's historical investment in information
technology research. In 1995, the National Research Council
documented the return on this investment. The study cited
numerous examples of information technologies whose roots lay
in federally funded research or that were nurtured through
critical development periods by Federal research. Examples
include network technology in the Internet, the Web browser,
computer windowing, computer graphics, reduced instruction set
computers, design of very large scale integrated circuits, data
storage technology, and parallel computing architecture.
In 1999, the National Research Council extended its 1995
conclusions, citing additional contributions to technology and
to the economy. Federal information technology research also
returns value directly to government operations through at
least two pathways, the first through government purchase of
commercial off-the-shelf information technology products that
have been invented or improved through Federal research. The
second pathway is through the development of special
information technology needed for government missions. This is
clearly shown in the government's research and development
programs, where many of the specialized information
technologies have been invented or developed by the Networking
and Information Technology R&D Program, often in direct
partnership with the program intending to use those
technologies.
Let me turn to my third point. This concerns the value of
current Federal investments in IT research. The Networking and
Information Technology R&D Program is currently working in
areas such as improving the quality and reliability of
software, improving the security of operating systems,
applications and networks, making it easier and more productive
for humans to interact with computer systems, including access
by individuals with disabilities, managing resources
distributed over the Internet, applying computer modeling and
simulation to scientific and engineering fields, detecting and
responding to natural or man-made threats, managing information
intensive dynamic systems and supporting lifelong learning.
Of perhaps special interest to this subcommittee is
research in information security. Federal agencies are funding
applied research to better enable us to cope with security
weaknesses in the architecture of operating systems, networks
and applications, as well as fundamental research,
investigating ways to improve the intrinsic security of these
architectures. The President's Information Technology Advisory
Committee is currently studying this area and will issue
recommendations regarding Federal research investments.
A specific example of the value of current Federal
investment concerns Google, and it may serve to illustrate the
value of this research generally. The Digital Libraries
Initiative is an ongoing part of the program that has been
sponsored by NSF, NASA, DARPA, and NIH. A recent article points
out that Google, the search engine company that is about to
issue a very significant initial public offering of stock, owes
its technology directly to a Digital Libraries Initiative grant
to Stanford University. Under this grant the cofounders of
Google invented, developed and tested their search algorithms.
My final point concerns the management of IT research.
Federal research programs have benefited from talented research
managers in the agencies and in funded projects. Because
research deals directly with the unknown and unanticipated, it
must be managed deftly. Often research failure becomes success,
as intractable obstacles point the way to alternative
approaches. Both Federal program managers and researchers must
have good instincts regarding when to continue the proposed
research and when to abandon or modify it.
Structures for managing and overseeing federally funded
research should allow program managers to alter projects in
midcourse in response to preliminary results and need to
recognize that research projects can produce valuable results
even if they do not achieve their original objectives. Failure
to manage deftly risks stifling creativity and innovation. The
history of information technology research demonstrates the
benefits of a flexible approach, and this approach is
consistent with the administration's R&D investment criteria.
This concludes my remarks. I thank the committee for the
opportunity to testify.
[The prepared statement of Dr. Nelson follows:]
[GRAPHIC] [TIFF OMITTED] 97999.006
[GRAPHIC] [TIFF OMITTED] 97999.007
[GRAPHIC] [TIFF OMITTED] 97999.008
[GRAPHIC] [TIFF OMITTED] 97999.009
[GRAPHIC] [TIFF OMITTED] 97999.010
[GRAPHIC] [TIFF OMITTED] 97999.011
[GRAPHIC] [TIFF OMITTED] 97999.012
[GRAPHIC] [TIFF OMITTED] 97999.013
Mr. Putnam. Thank you very much. I would ask the remaining
witnesses to please try to adhere to our 5-minute rule and
check the lights on the table.
Our next witness is Dr. Peter Freeman. Dr. Freeman is the
assistant director for the Computer and Information Science and
Engineering Directorate. He was previously at Georgia Institute
of Technology as professor and founding dean of the College of
Computing since 1990. From 1987 to 1989 he served as division
director for computer and computational research at the
National Science Foundation and helped to formulate the High
Performance Computing and Communications Initiative of the
Federal Government. In addition to his many activities as dean
at Georgia Tech, he headed an NSF-funded national study of the
IT worker shortage, started an active group for deans of IT and
computing, and published several papers relating to future
directions of the field. He received his Ph.D. in computer
science from Carnegie Mellon, his M.A. in mathematics and
psychology from UT Austin, and his B.S. in Physics from Rice.
His research and technical expertise has focused on software
systems and their creation.
We welcome you to the subcommittee. You are recognized.
Dr. Freeman. Thank you, Chairman Putnam, Ranking Member
Clay. Good afternoon. It is a pleasure to be here this
afternoon and to have the opportunity to testify before you and
to discuss information technology R&D. Let me begin by
clarifying some terms that I think we will all be using this
afternoon.
It is important to understand that the subject of today's
hearing, IT R&D, is open to multiple interpretations that can
lead to misunderstandings and to differences in reported
activity levels. For example, it is often reported that a
company spends a huge sum on IT R&D, but a closer examination
almost always reveals that the vast majority of that sum is
actually spent on development, not research. In the past, the
term ``information technology'' was usually taken to refer to
data processing activities such as payroll, accounting or
inventory, not the full range of work to which the term now
often refers. I would note that the Federal R&D community
primarily uses the more general meaning of the term
``information technology.''
Definitions of research and development are notoriously
overlapping and often lumped together. In the technical
community, research generally refers to activities that produce
new knowledge, while development refers to the use of existing
knowledge to produce new systems, products or practices. Even
these very general definitions are open to much interpretation
and practice. An important distinction, however, is that
research is usually targeted more broadly to longer term and
must be provided a very broad and loose type of oversight,
while development usually has very specific targets, has a
shorter timeframe, and requires a project management type of
oversight.
Let me now outline two frameworks for discussing Federal
activity in this area. The first separates IT from its usage.
Very simply, it is often useful to differentiate between IT
activity and IT-enabled activity. For example, a research
project we are currently supporting at NSF, an assessment of
voting technology and ballot design, seeks to provide an
assessment of information technologies relative to on-line
voting and ballot design. This is certainly IT research. It may
lead to some IT development of, for example, better e-voting
systems, but use of those systems would certainly be IT-enabled
activity.
The second framework that I would note is the one we use to
report Federal activity in this area. The major research
emphases of the NITRD effort are called program component
areas, and those are spelled out in what we call the Blue Book,
our annual supplement to the President's budgets.
Let me now turn to the questions expressed in your letter
of invitation. The first question was who is doing IT research
and development? I believe, as Dr. Nelson has already
indicated, at least 13 agencies or major subareas of larger
agencies report work in the NITRD program that is self-
identified as research. Non-U.S. Government personnel perform
the majority of that work, as Dr. Nelson has indicated. There
is undoubtedly additional IT research supported by the
government and of course a very large amount of development, as
Ranking Member Clay's opening statement made note of.
The second question was where are these investments being
made. I think it is fair to say that there is some amount of
investment in every State, in every research university and
essentially every company capable of providing research service
to the U.S. Government.
Your third question is what is government gaining from
these investments. In general, government is gaining directly
from the technical base used by our military and for
streamlined governmental operations and indirectly by fostering
the continuing economic revolution that provides the
innovation, productivity and economic vigor for our Nation as a
whole.
Your fourth question was why should government continue to
make those investments. I can only add that as industry often
and publicly stresses, it is because federally funded research
is essential to the continued advancement of IT technology.
Your final question, how much is being spent by the Federal
Government? I believe that within the stated caveats, the
cross-cuts listed in our annual Blue Book provide a good
compilation of Federal research activity in this area.
In conclusion, let me thank you for the opportunity to
appear before you today. I would ask that my fuller written
statement be entered into the record. I will be glad to respond
to your questions.
[The prepared statement of Dr. Freeman follows:]
[GRAPHIC] [TIFF OMITTED] 97999.014
[GRAPHIC] [TIFF OMITTED] 97999.015
[GRAPHIC] [TIFF OMITTED] 97999.016
[GRAPHIC] [TIFF OMITTED] 97999.017
[GRAPHIC] [TIFF OMITTED] 97999.018
[GRAPHIC] [TIFF OMITTED] 97999.019
Mr. Putnam. Thank you very much.
Our next witness is Dr. Hratch Semerjian, who is serving as
Acting Director of NIST. He has served as the Deputy Director
of NIST since July 2003. In this position Dr. Semerjian is
responsible for overall operation of the institute, including
financial management, human resource management facilities and
information technology systems, effectiveness of NIST technical
programs and for interactions with international organizations.
Dr. Semerjian received his Master's and Ph.D. Degrees in
engineering from Brown in 1977. He joined the National Bureau
of Standards, now known as NIST, where he served as director of
the Chemical Science and Technology Laboratory from April 1992
through July 2003. He has received countless awards and we
welcome him to the subcommittee today.
You are recognized for 5 minutes.
Dr. Semerjian. Thank you, Chairman Putnam and Ranking
Member Clay. Thank you for the opportunity to testify about
NIST's contributions to Federal information technology research
and development.
The impacts of information technology on the United States
and the world economy are certainly well known. NIST plays a
critical role in building trust and confidence in information
technologies and development of secure, reliable and
interoperable IT systems. NIST's programs help to ensure that
the U.S. industry maintains the competitive advantage vis-a-vis
the rest of the world while ensuring that U.S. Government
information technology assets remain secure.
Twenty-first century science is being pushed by continuing
progress in computing information and communication technology
and pulled by the expanding complexity, scope and scale of
today's technological challenges. Information technology is
providing the potential for the research community to build new
types of scientific and engineering knowledge and to pursue
research in new ways and with increased efficacy. The key to
these breakthroughs is achieving the necessary functionality,
interoperability, usability, confidence and data protection
within the IT systems that will lead the way.
NIST is at the forefront of these developments. I'd like to
give you just a few highlights from our information technology
program which will give you a flavor of the wide array of
expertise that exists at NIST. For example, NIST ensures the
security, confidentiality, integrity and availability of
information by providing standards and guidelines, testing
methodologies, and other Federal Information Processing
Standards [FIPS], in compliance with legislation such as the
Computer Security Act, the Federal Information Security
Management Act and the Cyber Security Research and Development
Act.
NIST helps mitigate the cost of inadequate software
testing, which is estimated to be around $60 billion, by
developing test methodologies for software assurance and
conformity to IT standards.
NIST develops tests and measurement technology that keys
the implementation, robust operation, and continuity of
operations of the Nation's core networking infrastructure,
especially to assure the robustness of the systems under
various failure and recovery scenarios.
NIST enables efficient access, manipulation, and exchange
of complex information through advances in human language
technology that enhances context extraction, question answering
and speech-to-text capabilities.
And NIST provides analytical, statistical and computational
tools for solving scientific and engineering problems. Some of
these tools, for example, are currently being used in the
analysis of the World Trade Center collapse. We are also
collaborating with the semiconductor industry to create a Web-
based electronic handbook of statistical methods.
Through these efforts, NIST has developed world class
competencies in cyber security, software, networks information
access, mathematics, statistics and interoperability. This
bundle of competencies, combined with Nobel prize winning
expertise in the physical sciences, places NIST in a unique
position to create an enormous impact on the economy and
innovation enterprise in the United States. It is precisely
this unique capability that attracts industry and other Federal
agencies to collaborate with NIST. And let me highlight some of
these specific efforts where NIST's expertise is being put to
direct use.
For example, NIST works with industry to ensure the
interoperability of technology specifications. Interoperability
is essential to productivity and competitiveness of many
industries because efficient design and manufacturing require
the coordination of many different participants and processes
that rely on a digital representation of the product. To
mitigate the billion dollar annual cost just to the automotive
supply chain, NIST has initiated the NIST manufacturing
business-to-business operability test beds, for example.
NIST assists government and industry in protection of the
U.S. borders through the development of biometrics evaluation
systems, standards and research. Two recent laws recognize this
expertise and provide specific requirements for NIST, the U.S.
Patriot Act and the Enhanced Border Security and Visa Reform
Act.
NIST enhances trust and confidence in voting systems. The
Help America Vote Act provides NIST with mandates in the areas
of security, hardware and software interoperability and human
factors issues. Under HAVA, NIST just recently released a study
on human factors which will be used to improve the performance
and reliability of voting machines.
And looking more into the future, NIST makes revolutionary
advances in quantum communications and computing. This is
really important for the future of the country, because quantum
communications offers the promise of perfectly protected
messages, while quantum computing offers the promise of
dramatically increased computing power.
NIST also utilizes information technology for knowledge
management. I think we have a knowledge-based economy, and both
the creation and dissemination of knowledge is a very important
part of what we do at NIST.
NIST works with other Federal agencies, academia and
industry to develop and promote openness and interoperability
of information technology. We work with other agencies to
provide expertise in our own unique areas of research to DOD,
DHS, HHS, DOJ and others. We also participate in the
Interagency Working Group on Information Technology R&D, and we
also cooperate with industry on integrating information-based
manufacturing systems and development of the measurement and
standards infrastructure needed for the application of
intelligence systems in manufacturing, defense and homeland
security.
In fiscal year 2004, the NIST Information Technology
Laboratory received about $48 million in appropriated funds. In
addition, ITL received about $17 million from other agencies on
a reimbursable basis. These are the words with M, not the B.
The President's 2005 budget request has an increase of $7
million for the NIST IT budget.
In conclusion, NIST takes its roles in maintaining the
vitality of the U.S. information technology industry seriously
in providing unique expertise to the rest of the government and
in sharing with industry, government and universities the basic
science and technology that comes from its measurement and
standards research. These brief examples of our work and
accomplishments illustrate NIST's commitment to these roles.
They also demonstrate the base upon which NIST continues to
build.
This concludes my prepared remarks. Thank you.
[The prepared statement of Dr. Semerjian follows:]
[GRAPHIC] [TIFF OMITTED] 97999.020
[GRAPHIC] [TIFF OMITTED] 97999.021
[GRAPHIC] [TIFF OMITTED] 97999.022
[GRAPHIC] [TIFF OMITTED] 97999.023
[GRAPHIC] [TIFF OMITTED] 97999.024
[GRAPHIC] [TIFF OMITTED] 97999.025
Mr. Putnam. Thank you very much.
Our final witness for this panel is Dr. Carl Edward Oliver.
Dr. Oliver is the Associate Director of the Office of Advanced
Computing Research for the Office of Advanced Science at the
Department of Energy. He is responsible for basic research and
applied mathematics, computer science and networking needs in
the Office of Science. His duties include management of the
Small Business Innovative Research Program.
Dr. Oliver came to DOE under the Intergovernmental
Personnel Act from Sandia National Laboratories. Prior to that
he was the associate laboratory director for computing robotics
and education at Oak Ridge National Lab from 1995 to 2000.
After receiving his Ph.D. in mathematics in 1969 as a NASA
fellow from the University of Alabama, he held research and
management positions at the Air Force Weapons Lab, the Air
Force Office of Scientific Research and DOE.
He has also held teaching positions at six universities,
been active on a national and international level organizing
professional society meetings for numerous academic societies
and has served on several university and Federal advisory
committees and others under the auspices of the OSTP and the
National Science Foundation.
Welcome to the subcommittee. You are recognized for 5
minutes.
Dr. Oliver. Mr. Chairman, I too commend you for holding
this hearing and I appreciate the opportunity to testify on
behalf of the Department of Energy's Office of Science on a
matter of importance to the Nation; namely, information
technology research.
Dr. David Nelson and Dr. Freeman have given you some
overview of the Federal IT R&D activities. I'll concentrate on
those areas of the portfolio where the Office of Science
focuses its efforts: High performance computing, large scale
networks and software that enables scientists to use these
resources as tools for scientific discovery.
Ever since the inception as part of the Atomic Energy
Commission immediately following World War II, the Office of
Science has blended cutting edge research and innovative
problem solving to keep the United States at the forefront of
scientific discovery. Since the 1940's, the Office of Science
supported the work of more than 40 Nobel prize winners.
Research supported by the office has made major contributions
to the United States in research areas such as magnetic
resonance imaging, medical isotopes, composite materials used
in motor vehicles and x-ray diagnostic of computer chips and
other high tech materials.
Other research investments have led to such innovations as
the Nobel prize winning discovery of new forms of carbon,
noninvasive detection of cancers and other diseases, improved
computer models for understanding global climate change and new
insights into the fundamental nature of matter and energy.
High end computing has become an indispensable tool for
researchers across the Office of Science. Large multi-
disciplinary teams of researchers that combine the expertise of
physicists, chemists or biologists with the expertise of
computer scientists and mathematicians are working on the next
generation of computational science tools that will enable the
discovery and design or advance of materials for the
development of catalysts that dramatically reduce the energy
costs and emissions and understanding of the dynamics of
combustion systems. Each of these examples and many more will
have a significant effect on the missions of the Department of
Energy and then the missions of other U.S. Government agencies.
High performance networks play a critical role as well
because they make it possible to overcome the geographical
distances that often hinder science by making all the
scientific resources readily available to scientists,
regardless of their physical location. In this area, we work in
close coordination with the National Science Foundation and
university consortia such as Internet II to ensure that
scientists at universities can seamlessly access unique DOE
facilities and their scientific partners in DOE laboratories.
To develop these tools we also work closely with other
agencies. A significant part of the coordination has been
described by Dr. Nelson already. In high end computing we
cochaired the High End Computing Revitalization Task Force that
was put together. This task force identified our Nation's
critical needs in a report released in May, and they proposed a
game plan to improve U.S. computing capabilities.
The Office of Science and other Federal agencies are
working to implement the recommendations of the task force
report and to develop the next generation of supercomputing
capability as well as networks needed to allow the broadest
possible access to new systems.
On May the 12th of this year, Secretary Spencer Abraham
announced that the Department of Energy will provide $25
million in this fiscal year to a team led by Oak Ridge National
Laboratory to begin to build a new supercomputer for scientific
research. This is an important step toward achieving our
leadership goals. When complete, researchers will gain the
ability to understand the natural world with the precision that
could only be imagined a few years ago.
It's clear that working with our computing industry, we can
build these tools. The administration has developed a clear
path forward for revitalizing high end computing, and with
vital support from the Congress and administration I am
confident we will succeed.
Once again, thank you for the opportunity to testify before
the committee on this important matter.
[The prepared statement of Dr. Oliver follows:]
[GRAPHIC] [TIFF OMITTED] 97999.027
[GRAPHIC] [TIFF OMITTED] 97999.028
[GRAPHIC] [TIFF OMITTED] 97999.029
[GRAPHIC] [TIFF OMITTED] 97999.030
[GRAPHIC] [TIFF OMITTED] 97999.031
[GRAPHIC] [TIFF OMITTED] 97999.032
Mr. Putnam. Thank you very much, Dr. Oliver. I want to
thank all of you for your opening statements. Your full
statement is to be included in the record, and we will begin
with questioning.
Among these two dozen or so agencies that are coordinating
or, excuse me, are engaged in Federal research, is there some
collaboration across those agencies to pursue an overall
strategic framework for research? Do they communicate with one
another to avoid duplication or to buildupon the successes that
are being found? Does the left hand know what the right hand is
doing?
Dr. Freeman, you're nodding. We'll let you answer first.
Dr. Freeman. That's fine. Thank you sir, and I think that
my cochair of the NITRD working group, David Nelson, will add
to whatever I may have to say.
I think the answer to your question is a very definite yes
in terms of coordination, in terms of knowing what each other
is doing. In some cases that results in joint solicitations.
For example, in the high performance computing area, we are
currently coordinating, or I should say collaborating with
DARPA for a jointly funded research program. We have several
others in other areas already in place with DARPA. We are
putting one in place with NIH, so there are a number of
examples of that sort.
The Interagency Working Group as a whole meets every 3
months at which representatives from all 13 of those agencies
are present to review budgets, to review strategic directions
and, in general, to coordinate. There are then the program
coordination areas that both Dr. Nelson and I mentioned. Those
smaller, more narrowly defined groups; for example, in the area
of human computer interface or in the area of high confidence
or secure systems, those smaller groups of program directors
and program managers typically meet on a monthly basis, in some
cases even on a weekly basis, because many of the primary
agencies are located physically adjacent to each other.
So there is a high degree of communication and
collaboration there.
Mr. Putnam. Dr. Nelson.
Dr. Nelson. I agree with the comments of Dr. Freeman and
would just add a couple of things. First, I refer to the
Interagency Working Group, which Dr. Freeman and I cochair, as
our board of directors. It sets the general directions and
identifies special topics. About a year ago, Peter and I
commissioned a study of whether the program components areas
and the groups that implement them were properly constituted,
and the answer that came back from them, and the Interagency
Working Group has blessed it, is that in general they are, but
in particular we need to focus more on security. And so we have
started as a cross-cutting topic looking at how security
aspects of all of our work need to be better done. It's, as you
know, an extremely difficult topic.
The other thing that I would add is that there's always an
open door to the involvement by additional agencies and
additional programs in the work of the--and I will use the
acronym here--the NITRD program. So for example, we have the
Federal Aviation Administration, Food and Drug Administration
as observers. They don't have much research, but they do want
to use what the research program produces. And we also are
discussing with the Department of Homeland Security its entry
in the program. So the door is always open. Information
Technology Research and Development has an awful lot of small
pieces. And so as Peter has said, much of the detailed work is
done putting those pieces together.
I mentioned before an example of research needs
assessments. That's usually the first step. That's one. Another
example of research needs assessment is the document that Peter
just referred to, which is the Federal Plan for High-End
Computing, which involved about 60 Federal managers from 12
agencies for about a year. After research needs assessments is
research planning. Peter gave one example. There are many
others.
The agencies share reviewers: They coordinate the
solicitations; and then they jointly review progress and, where
necessary, make changes and of course then go through another
cycle of research needs, research planning, research
implementation, research review.
Mr. Putnam. OK. Dr. Freeman, you said that they meet
periodically to review strategic direction. Who sets the
strategic objective, and could you summarize what the strategic
objective is currently for Federal IT research and development.
Dr. Freeman. I don't think that I or anyone is really
capable of saying what the objective is. As Dr. Nelson just
indicated, there is a fairly elaborate process of trying to
understand the needs for future research. He mentioned some of
those activities. Let me share another one with you that
perhaps will illustrate the process and thereby how the overall
objectives are ultimately set.
Another activity that Dr. Nelson and I initiated, oh,
probably a year and a half ago now, was an effort by the
members of our working group to look at what are the major
challenges, often called grand challenges, that would involve
information technology, not on the research but at the usage
end; for example, being able to seamlessly access the medical
records of any citizen anywhere with appropriate privacy
security, etc. It was put forth as a possible grand challenge,
because we don't really know how to do that today. That working
group, which worked for, oh, 6, 8 months, ultimately came back
with a set of these grand challenges that, if achieved, would
have great benefit for our society and for our economy. Taking
those grand challenges, they then backed up and said, so what
research should we be doing? So for example, security was
something that was seen to be a critical component of
essentially every one of those grand challenges.
So, through this process, our members, the various
agencies, and through other processes that are not a part of
the interagency activity, because each agency has its own
internal processes to bubble up these research needs,
prioritization then of those research objectives will depend
upon the individual agencies, upon their missions, and upon
their judgments as to which of, as Dr. Nelson has indicated,
the many, many objectives that one could name.
So, in sum, what we wind up with is not a single strategic
objective or even a coherent set of a small number. Obviously,
we can boil those up to a high level and certainly security is
a current high objective, I believe, of all of our members.
High performance computing is another objective that is very
important to many of the agencies. Ease of access to large data
stores is a third objective. But I would not characterize those
as forming a strategic plan in the same way that an individual
company or an individual agency might have.
Mr. Putnam. Dr. Oliver, a lot of attention has been paid to
the fact that Japan now has the fastest supercomputer, the
Earth simulator, which I believe has led the United States to
reassess its high end computing R&D plans. Given where we now
stand, are the existing Federal efforts appropriately targeted
to deal with the challenge of positioning the United States as
a leader in IT R&D, or are we losing our leadership position in
this area?
Dr. Oliver. I think that the formation of the High End
Computing Revitalization Task Force last summer was the right
step at the right time for us to articulate as a group of
Federal agencies and departments how to address this problem.
There were lots of meetings that were held and the report was
written, and I think the implementation plan is pretty much in
place or will be soon. So I think fully supporting that plan
will lead us to a leadership position again. There is no doubt
about it.
Mr. Putnam. Mr. Clay.
Mr. Clay. Thank you, Mr. Chairman. Dr. Nelson, according to
the recent report on Federal R&D released by RAND, total R&D
funding for colleges and universities grew from approximately
$70 billion in fiscal year 1996 to $96 billion in fiscal year
2004. Yet only $2.2 billion is dedicated specifically to IT
research and development through the NITRD program.
Can you tell us why it is such a small percentage of the
total allocation, and have no concerns with cyber security and
national defense caused your office to reevaluate its support
for IT R&D funding?
Dr. Nelson. Yes, Congressman Clay, the priorities for
research, as for any other Federal program, go through many
steps and have many masters. And so addressing a question like
why IT R&D is only $2 billion is almost impossible. One could
equally ask why is it as big as it is. And I think, as you
pointed out, it has grown substantially in recognition of its
importance.
I would suggest though, and I think Dr. Freeman has also
mentioned this, that much of other research has information
technology components. The term that's used often is it's
embedded in that research, and it is almost impossible to tease
out how much of that other research is information technology.
But we know there is a lot there. In other words, the amount of
information technology research is higher than the numbers
might suggest.
The second thing that I would say is I believe the RAND
study pointed out that much of that increase went to medical
schools and was a direct result of the doubling of the National
Institutes of Health budgets, and of course that has been a
national and bipartisan priority. It is good to note that much
of the improvement in health care research has come about from
better usage of information technology, and bioinformatics is
now a thriving field.
So again it's hard to tease out, and the bottom line I can
give you is that there probably isn't a direct answer. We can
all hope that those good research ideas are funded and that
they appropriately impact our economy and way of life.
But the process for arriving at it is probably beyond that
of any one person to comprehend.
Mr. Clay. Then another question, Doctor. Tell me about,
compare the 2004 budget for Federal IT R&D to 2005. It's been
reduced by about $200 million, from $2.2 billion to $2.0. Can
you tell us why the amount requested was reduced?
Dr. Nelson. Yes. I would like, if I could, to answer that
for the record. And I will tell you why. We have a total for
2005, but because the--and I will use the term NITR&D, that's
Networking and Information Technology R&D--program is what we
call a crosscut, you can't get there by aggregating line items.
And, therefore, it takes judgment and allocation to come up
with either the total or with the subcomponents. We do not yet
have the 2005 numbers for those subcomponents.
And before I could address why that $200 million is
approximately about 10 percent down, I would have to have a
better idea of how it matches up with the program component
areas that Dr. Freeman referred to, and therefore what the
change is. I would be happy to respond for the record, but it
may take a few weeks before we get back to you.
Mr. Clay. That's fine. Thank you for that response.
And Dr. Freeman, I am concerned that a large proportion of
Federal R&D funds are being limited to the life sciences. Can
you tell us if the IT needs of the medical and biological
research communities are being met through the current formula
for Federal R&D funding?
Dr. Freeman. Congressman Clay, I'm afraid in the specifics,
I am not competent to answer that question. Obviously, someone
from NIH or the medical community should address it. I would
note that through the interagency working group and our budget
reviews, we are aware that NIH is spending large sums on
information technology usage, more on the development and usage
side of the ledger than on the underlying research. But I will
have to defer to their judgment as to whether there is
sufficient IT research being done in the country to support
those medical efforts.
Mr. Clay. OK. Then I guess asked another way. Can you cite
for us what other Federal R&D needs would rival the need for
life science research?
Dr. Freeman. Well, certainly based on the demand through
the proposal submission process at NSF, there is a very large
unmet need for additional research. Let's take the cyber
security area. My program director in this area is just in the
process this week and next of making final decisions on
something like 150 proposals that were submitted by a deadline
a few weeks ago for funding specifically in the cyber security
area. I have not seen the final results, but based on previous
competitions, probably at least 30 to 35 percent, about a third
of those proposals would be ranked by their scientific peers as
worthy of funding, that is, scientifically valid. We will be
lucky if we can fund 10 percent of those submitted proposals.
So there is an overhang there of good ideas in the cyber
security area that we are simply not going to be able to fund.
I suspect that my colleague from NIST has a similar experience
in that area.
Mr. Clay. I will get to him. Thank you for your response.
You know, you guys have long responses and long opening
statements, and they give us a limited amount of time. But I
appreciate your effort.
Dr. Oliver, in what ways can the Office of Science
collaborate with the private sector on R&D initiatives in order
to find solutions to problems facing our Nation and the
international community, such as environment and educational
issues?
Dr. Oliver. Well, we have had a long history of
collaborating with the computing vendor industry in high-end
computing and network and so forth. And, you know, we have
cooperative research and development agreements with many of
the individuals who have funding from us, especially in the DOE
laboratories. And so that has tremendous benefit in many ways
with the computing industry. And the wave goes out from there.
We get in machines. We have often bought serial one of every
high-end machine in the Department of Energy from the beginning
of time. There are a few exceptions.
So we get them when they are raw. They don't necessarily
work. We invest in operating systems to help make this work. We
do this in cooperation with other agencies, DARPA and 1-Ks or
NSF in another. And so we help make a machine viable. We then
put on applications software like a structural analysis code
and make it work on the machine. That makes it attractive to
the aerospace companies and so forth.
So there is a dramatic effect like that where we work with
industry, you know, to help in the economic sense. And, but we
do it to serve our own needs. I mean, we have to make these
machines work in order to meet the missions of the Office of
Science to do science.
Mr. Clay. Thank you. And, finally, Dr. Semerjian. Are there
specific areas of IT R&D that the government is currently not
pursuing but merit consideration in future strategic plans?
Dr. Semerjian. Well, that is a very broad question.
Probably some of my colleagues are better prepared to answer
that. But just from our own point of view, the kind of things
that we do at NIST, I think I agree with Dr. Freeman that cyber
security issues I think really need to be addressed. Because in
every application area that we look at, whether it's the
medical records or whether it's communication issues and
national security issues, cyber security continues to be at the
top of the list of issues, which is clearly underfunded.
And I think we are very pleased that the President's budget
proposes to increase the NIST part of that. But that is only a
drop in the bucket. I am sure there are many other security-
related issues in other agencies which need to be addressed.
Mr. Clay. Thank you for your response. I thank the panel
for their indulgence. Thank you, Mr. Chairman.
Mr. Putnam. Thank you, Mr. Clay. I just want to followup a
bit on the coordination line of questioning. To what degree are
the classified and unclassified research initiatives
coordinated? If you take cyber security, for example, it's
certainly broader than NIST and NSF. You have an alphabet soup
of agencies within the intelligence community who are also
doing work. Do they share some of their direction on research
as well, or is that a world unto itself?
Dr. Nelson.
Dr. Nelson. I can speak to the relationship between the
classified world and the unclassified. I personally have high
security clearances, but the NITRD program is totally
unclassified. Now, I believe there is reasonably good
coordination between the classified side and the unclassified.
And I will say how that occurs. Several of the agencies that
participate in the NITRD program, in particular the National
Security Agency, the National Nuclear Security Agency, and the
Defense Advanced Research Projects Agency, bring their
unclassified research to the table and work closely with
agencies like NSF, NIST, DOE science, and so on to get the best
bang for the buck out of that. Now, clearly, they take those
results back into the classified side, and the NITRD program
does not see those.
I am aware that there are coordinating groups that deal
with classified research, in particular, the Infosec Research
Council. But I do not know whether that is broadly the case. I
could mention another agency or two that interacts with us on
an informal basis, the Central Intelligence Agency, and as I
said earlier, we are talking seriously with the Department of
Homeland Security about their joining the program. So, indeed,
there is coordination. As it comes through the boundary between
classified and unclassified, I'm less knowledgeable about what
goes on behind the classification screen.
Mr. Putnam. So Homeland Security is not currently a part of
this NITRD program?
Dr. Nelson. That's correct.
Mr. Putnam. And what is it that needs to be decided to
allow them to participate?
Dr. Nelson. As you know, they have been forming up. And
bringing all those subagencies together has been a monumental
task. So the short answer is they are still working on their
research agenda. They are still getting staffed up to carry out
that research agenda. And I fully expect that as soon as they
have people who are able to work--coordination takes time, it
takes people--that they will join.
Mr. Putnam. Well, I certainly expect them to get ramped up
and be a part of it as soon as possible. They are an important
piece, and they are a relatively new agency, but they are no
longer a new agency and that excuse only carries you so long.
Dr. Nelson. Yes, sir.
Dr. Semerjian. If I could add on the classified versus
unclassified issue. We do have a close working relationship
with NSA. It is formally recognized that NIST provides the
standards for the unclassified world, so to speak, and NSA for
the classified. And we do work closely with them, since some of
the technologies are utilized on both sides of the fence. So at
least in terms of standard issues, cyber security-related
standards, we have a very good working relationship with NSA.
Mr. Putnam. Good. I want to give all of you the
opportunity, because we were cut short on the opening
statements, but we have another panel that we need to move to.
I want to give any or all of you, if you have something that
you wish to add to this conversation, something you wish you
had been asked, now is the time. So we will begin with Dr.
Oliver, work backward, and then we will seat the second panel.
Dr. Oliver, anything to add?
Dr. Oliver. Nothing to add.
Dr. Semerjian. (Shaking head.)
Mr. Putnam. Dr. Freeman.
Dr. Freeman. A former professor always has another word.
One topic we have not discussed today is the issue of
education. That is one that, of course, NSF has a key
responsibility in. And I would note that it is also a key
element in both the research and certainly the development
aspects of information technology, to say nothing of the
utilization of it. I think that your opening statement,
Congressman Clay's opening statement very aptly recognize the
importance of information technology in essentially every
element of our life today. And I would only add that we must
keep in mind that educating all of our citizens at all levels
of the work force, all the way down to kindergartners, in the
usage of that technology is a key challenge that we need to
keep in sight.
Mr. Putnam. Let me followup on that and ask you, as someone
who has come out of the academic world. Many of us are familiar
with and very concerned about the decline in math and science
skills among America's young people, the number of degrees
being awarded to Americans in a number of these sensitive
fields. I am not asking you for a silver bullet, but what are
the steps that we can begin to take to turn that around and
produce more math and science graduates?
Dr. Freeman. I'm glad you didn't ask for the silver bullet,
because I certainly don't have it and I'm not sure that I have
all of the steps. But it is something, and the reason I bring
it up here is it is a topic that I believe we must first and
foremost always keep in mind. So it is important, of course, to
look at funding levels at the substance of the research, but we
must be mindful that we have to have the educated people to
carry out that research, for example. So I think the attention
of committees such as yours to the educational issues, the
attention of all of us as citizens is certainly a first and a
very necessary step to take.
Mr. Putnam. Dr. Oliver, I believe in some of your
responsibilities at the Department of Energy, do you have a
hard time filling slots in the Office of Science? Do you have a
hard time recruiting good people who want to work for the
government doing this type of research?
Dr. Oliver. You have done your homework. Yes, it is, in
many areas, very difficult to get program managers to come and
work in D.C., though I tout the city as the greatest place to
live in the country in opportunities and everything. I mean, I
think we have a terrific, you know, organization. And the jobs
are truly exciting. And I find that it's just getting more and
more difficult in our area and I don't know why.
Maybe industry salaries are high, academic salaries are
very high for people in computer science, applied math, that
know about high-end computing, in tremendous demand, and I
think maybe the pool is a little small. I mean, it's the supply
and-demand situation. Anyway, it's very difficult for us to get
people in the Office of Science. And we have a very important
challenge facing us because I think we are an aging group, not
just in information technology, but throughout the program
management staff. And it's something we are aware of and that
we are trying to address, and we are looking at all of the
things that you can do, knobs you can turn. But it is indeed a
challenge. So we have a lot of people with a lot of dual jobs,
dual-hatted, but they rise to the occasion.
Mr. Putnam. Thank you.
Dr. Nelson, final comment?
Dr. Nelson. Yes. Just one footnote in the planning and
management area. We can always do a better job of strategic
planning, no question of that. But we have to remember that in
research, it's often the least planned and the least unexpected
that yields the biggest dividends. Who would have thought that
the high performance computing program would produce the Mosaic
browser, and yet it was the NSF program at the University of
Illinois that did just that. Who would have thought that an
early program to link academic computers would have produced
the Internet? But it did just that. And so in research we
always look for the revolutionary change, the things that
really improve the economy in the country. It's hard to plan
those.
And, as I mentioned in my testimony, management of those
activities has to be very deft and sensitive. It's very easy to
stamp out the revolutionary and it's very hard to encourage it.
Mr. Putnam. Thank you very much. And I want to thank the
entire panel. And, with that, the subcommittee will stand in
recess while we arrange for panel two. And if you would be
seated as soon as possible. Thank you, gentlemen, very much for
your contributions to this hearing.
If the second panel could please take their seats, we will
begin. The subcommittee will reconvene. And if you would please
rise, and anyone accompanying you who will assist you in
answering the questions please rise for the administration of
the oath.
[Witnesses sworn.]
Mr. Putnam. Note for the record that all of our witnesses
responded in the affirmative. And we will move to the
testimony. You will note the lights on your table. Green light
means talk away. Yellow light means bring it in for a landing.
Red light means bring it to a close.
Our first witness is Dr. Donna Fossum. Dr. Fossum is a
senior scientist and legal policy analyst in the RAND Resource
Management Department, and is program manager of the RaDiUS
Project. Prior to joining RAND, Dr. Fossum served as the legal
counsel and technology specialist of the Committee of
Government Operations here in the House of Representatives. Dr.
Fossum has also served as the deputy associate administrator of
the Office of Federal Procurement Policy in OMB, where she
devoted much of her time to advising the administrator for
Federal procurement policy on matters involving the defense
industrial base. Most recently, she served as the senior
adviser for science resources development at NSF.
Dr. Fossum's work for the institute has centered on
developing a comprehensive data base of the R&D activities
sponsored by the Federal Government known as RaDiUS, to
facilitate the management and content assessment of the Federal
R&D portfolio. Much of her time is also devoted to working with
OSTP and other Federal agencies as well as numerous nonFederal
entities to identify and evaluate Federal activities in every
conceivable field of R&D. Welcome to the subcommittee. You are
recognized for 5 minutes.
STATEMENTS OF DONNA FOSSUM, MANAGER, RADIUS PROJECT, RAND
CORP.; EDWARD LAZOWSKA, CO-CHAIR, PRESIDENT'S INFORMATION
TECHNOLOGY ADVISORY COMMITTEE AND CHAIR, DEPARTMENT OF COMPUTER
SCIENCE AND ENGINEERING, UNIVERSITY OF WASHINGTON; WILLIAM
SCHERLIS, PROFESSOR, SCHOOL OF COMPUTER SCIENCE AT CARNEGIE
MELLON; AND STEPHEN SQUIRES, CHIEF SCIENCE OFFICER, VICE
PRESIDENT, HEWLETT-PACKARD
Dr. Fossum. Thank you. Thank you very much. I have
submitted a long statement for the record, so I will make it
very short and summarize it. First I want to express my sincere
appreciation for the hearing today, because this is a topic
that often gets overlooked and, yet it is at the heart of
everybody's life and becoming more so.
Let me explain a little bit about the RaDiUS program and
why it even came to be. In 1992, we were supporting the work of
the White House in science and technology, and discovered that
the data that everybody wanted or needed to find out what was
going on in the world of Federal R&D and on various topics, we
didn't have at the right level of granularity. It was at the
program level, which wasn't adequate to answer their questions.
For instance, they wanted to know what was going on in
electronics, preventing violence in youth, automotive-related
technologies, global positioning, aviation safety, etc.
And this is where RaDiUS was born. RaDiUS systemically
tracks all the R&D dollars that are identified by Federal
agencies as being R&D and tracks them through the layers of the
bureaucracy down to where they are actually spent--where I say
``the rubber hits the road.'' And this is where the R&D is
actually conducted. Courtesy of the RaDiUS data system, we now
have a capability that is used extensively by many of the
agencies in the Federal Government many Federal contractors,
universities, and others all over the world to learn what R&D
is being supported by the USG. Actually we have discovered
there is very little duplication of R&D in the Federal
Government, but there are many, many opportunities where
leveraging of Federal R&D dollars is not happening.
Agencies in the same red area don't even know it, but
through RaDiUS, they can find out they are working in the same
field and can leverage their dollars more effectively.
RaDiUS, in 1998, was declared a ``best practice'' of the
U.S. Government by the General Accounting Office. Courtesy of
RaDiUS, RAND has been able to produce two reports. ``Discovery
& Innovation'' has been called a chest crusher by one of our
colleagues. What it provides is the first compendium ever of
all the Federal R&D activities by State and city in the
country. Where is it happening and what are they doing? This
was done in 2000. It's time to be updated but a whole lot of it
has not changed all that much. Where it's happening has not
changed.
I will leave a copy of it with you. And already today this
report that came out a couple of months ago has been cited,
``Vital Assets.'' It contains the first really accurate
assessment of where all the Federal dollars are going to
universities around the country. This kind of capability has
been developed, courtesy of RaDiUS.
But that's not what the topic of today's hearing is. You
have gone through a wonderful list of questions to which we
would all like the answers, but the one question that has been
hinted at briefly by some of the people here that I want to
focus on, which is a pivotal question in this area, is ``What
is IT R&D?''--because IT itself has evolved so tremendously in
the last 30 to 40 years, and the R&D associated with it has
evolved right with it. For instance, initially IT was just
physical components. We were building hardware. We were looking
at the mathematical equations that could be used for pulling
data together and manipulating it. At the beginning, that was
IT R&D--physical components. And there's still a lot of that
R&D work going on. As it evolved over time, IT R&D got into new
applications and the new infrastructures. This is where you
started spawning software and other applications like data
bases. Once you put it together with infrastructure,
applications, design, and development, what you have are all
kinds of ramifications and potentials that had never been
thought of in the very early days.
So what I've provided in my written testimony is a template
of essentially four definitions of IT R&D. One is very narrow,
which is physical components. And another is quite broad, as it
includes information technology functionalities, information
technology applications and infrastructure, and the
capabilities enabled by information technology. And since we
are the data people, we wanted to give you some hard numbers to
grab on to. So what is the Federal Government spending in IT
R&D?
If you take a narrow definition of IT R&D as physical
components, and run it through the RaDiUS data system, we come
up with the fact that there was about $1.5 billion of IT R&D.
And this is a very conservative estimate. This is for just the
physical components. If you broaden the definition to include
everything, all four aspects, again, very conservatively, it's
over $11 billion. That's about 12 percent of the Federal R&D
budget. By the way, we are only talking about the ``Conduct of
R&D'' when we talk about this information. Keep in mind that
one of the biggest sources of confusion when we talk about R&D
numbers in the Federal Government is often they include ``R&D
facilities'' as well as ``R&D equipment.'' That doesn't get you
the ``conduct of R&D.'' So in RaDiUS we only focus on the
actual conduct of R&D work. That's what these numbers are.
So you have a narrow definition and a broad one.
What we cannot tell you is whether this is valuable
research. We can tell you what it is and if it is meeting
objectives of the Federal agencies. We can give you information
on it. We can tell you what's going on. Let me give you a hint
about some of this. Keep in mind, the major agency doing R&D in
IT is DOD by far. In the narrow definition, the No. 2 agency is
DOE. In the broad definition, it's HHS. You have all kinds of
players here. NASA is a major player. You have NSF, and the
Department of Agriculture. Every part of the Federal R&D
community is working somehow in IT-related activities.
Let me give you a little idea of who is doing what.
Mr. Putnam. If you would, if you could summarize it. We
have a vote at 3:15, and I want to get through testimony before
they ring the bells.
Dr. Fossum. OK. You've got it. I will just submit this to
the record. But let's suffice it to say that people are working
at DOD on everything from dealing with strike aircraft to
virtual battlefields. At HHS, they are doing R&D on wireless
EKG chips. They are doing all kinds of R&D all over the
government dealing with IT. And we are here as the tool that
can tell you what's going on where and help to better
coordinate it. Thank you so much.
[The prepared statement of Dr. Fossum follows:]
[GRAPHIC] [TIFF OMITTED] 97999.033
[GRAPHIC] [TIFF OMITTED] 97999.034
[GRAPHIC] [TIFF OMITTED] 97999.035
[GRAPHIC] [TIFF OMITTED] 97999.036
[GRAPHIC] [TIFF OMITTED] 97999.037
[GRAPHIC] [TIFF OMITTED] 97999.038
[GRAPHIC] [TIFF OMITTED] 97999.039
[GRAPHIC] [TIFF OMITTED] 97999.040
[GRAPHIC] [TIFF OMITTED] 97999.041
[GRAPHIC] [TIFF OMITTED] 97999.042
[GRAPHIC] [TIFF OMITTED] 97999.043
[GRAPHIC] [TIFF OMITTED] 97999.044
Mr. Putnam. Thank you very much.
Our next witness is Dr. Edward Lazowska. Dr. Lazowska is
professor and Chair of the Department of Computer Science and
engineering at the University of Washington. He received his
Ph.D. from the University of Toronto in 1977. He has been at
the University of Washington since that time. His research
concerns the design and analysis of distributed and parallel
computer systems. He is a member of the NSF, CISE advisory
committee, Chair of the Computing Research Association, and
member of DARPA ISAT, and a member of the technical advisory
board for Microsoft research.
He is a member of the NRC's computer, science, and telecom
board, and served on the CSTB committee that produced evolving
the high performance computing and communications initiative to
support the Nation's information infrastructure. He is a member
of the National Academy of Engineering and a fellow of the ACM
and of the high Triple E. He is a leader in the Learning
Federation, a group that is concerned with using information
technology to improve learning at the college level. Welcome to
the subcommittee. You are recognized for 5 minutes. I would ask
all of you to please hold tight on the 5, because there will be
a vote at 3:15. Thank you.
Dr. Lazowska. Thank you very much, Mr. Putnam, and the
other members of your subcommittee. It is a pleasure to be here
to testify today.
As you said in your introductory remarks, much of the gains
in productivity in the U.S. economy over the past decade, the
really unprecedented gains throughout the 1990's, have been
shown to be due to efficiencies produced through information
technology. And IT and its advances are driving advances in all
fields of science and engineering. So what your subcommittee is
asking is, how does that happen? And the abstract answer is
it's a complex ecosystem that involves companies and
universities and the Federal Government. It's been working for
50 years. The United States is the world leader in innovation
in information technology today because of some formula that
none of us can quite get our hands around but that
fundamentally seems to work. So it has been a 50-year story of
success.
Every aspect of IT that we rely on today, every billion
dollar subindustry, traces part of its origins to the federally
funded university-based research program. You have a two-page
handout of my remarks today. And on the second page is a little
eye testing graph, which I won't try to describe to you now,
but this is from a National Academy study that Dr. Nelson
referred to. And what it shows is two dozen different billion
dollar subcategories of the IT industry. And for each one, it
shows the complimentary roles of university research funded by
the Federal Government, industrial R&D, and product
development, becoming a billion dollar industry.
So the interplays are very complex, and there have been
authoritative studies of this. But the key thing for you to
understand is the role that the Federal research program has
played in all of these technologies.
I am fond of saying if you want to do E-commerce, you have
to have a Internet, you've got to have Web browsers, you've got
to have high performance data base systems, graphical user
interfaces, public key cryptography for secure credit card
transactions. All of those are results of the federally
sponsored research program.
In planning policy--and a point that Dr. Freeman made very
clearly--it is important not to confuse industry R&D with
research that's looking 5 or 10 or 15 years out. And here is
just a concrete example. I'm on the technical advisory board
for Microsoft Research. Microsoft advertises that it will spend
$6.8 billion on R&D this year. OK. Of that $6.8 billion, only a
couple hundred million is Microsoft Research, the organization
that I advise.
$6.6 billion is engineering the next release of Word and
Excel and Power Point and Windows. This is really important.
But that's done by taking ideas out of the R&D larder and
putting them into products. It's engineering the next
generation of the product.
Now, it sounds like I'm castigating Microsoft, but I'm not.
In fact, on the contrary, Microsoft invests about 5 percent of
its R&D budget in activities looking out five or 10 or 15
years. Dell, Oracle, Cisco invest essentially nothing looking
more than one product cycle out. HP has a representative on
this panel and HP does look more than one product cycle out in
a style not unlike Microsoft. So does Intel. But many of the
major IT companies don't look more than one product cycle out
at all. And that's what defines our future, what makes sure
that we are going to be a leader 5 and 10 and 15 years out.
You heard from Dr. Freeman that another important
characteristic of the Federal research program is that it
produces people. A second graph on my handout is the Department
of Commerce work force projections for the next 10 years for
various fields of science and engineering. And what it shows is
a huge work force gap in information technology compared to any
other field of science and engineering.
The Department of Commerce projects that more than three
quarters of all the jobs that will have to be filled in all of
science and engineering in the next 10 years are IT jobs.
Recent increases in support for IT research have been
important but have fallen far short of the levels recommended
by the President's Information Technology Advisory Committee.
Mr. Clay observed a disconnect in funding choices made by
agencies. The third graph in my materials shows the increase in
the government supported R&D budget over the past 30 years. And
you see that the vast majority of that is increases in the
National Institutes of Health. Every other field is essentially
flatlined, although IT R&D has doubled over that period, you
can't see the increase on a scale that includes health and
human services.
My response to your question asked of another witness about
HHS IT R&D is that largely what they do is take innovations
that DARPA and the National Science Foundation have funded and
apply them to biomedical problems, as opposed to investing in
fundamental IT research. There are exceptions, you will hear
about the NIH bioinformatics program, but that's a few tens of
millions of dollars new this year out of a $30-plus billion
budget.
So one other point that I would make is there are, as you
heard, 13 or more Federal agencies investing in IT R&D. I think
if you look at the history of innovation, it is NSF and DARPA,
with work from Energy in high performance computing, that have
driven the lion's share of the innovations.
A couple more points and then I will conclude. The research
community has concerns with the low level of funding for the
NSF Computer and Information Science and Engineering
directorate. That budget has gone up in recent years, but the
research budget there is still only a bit more than $400
million. We have concerns about----
Mr. Putnam. If you could bring it in for a landing, please,
sir.
Dr. Lazowska. Sure. We have concerns about DHS's failure to
invest in cyber security R&D. DHS began a year ago with a new
$800 million research budget and proposed allocating $7 million
of that to cyber security. That is simply a failure to
understand the threat posed by cyber terrorism. And it's not
that E-Bay goes down so you and I can't buy stuff; it is that
computers are in the control loop of every element of the
Nation's critical infrastructure. So if you want to attack the
electric power grid, you go after the control systems. So it's
a serious issue.
Summary. The track record is clear, the Federal R&D
investment has stimulated America's world leadership, our
economic boom, our boom in all science and engineering. Current
levels of Federal investment in IT R&D continue to be
dangerously low. Thank you.
[The prepared statement of Dr. Lazowska follows:]
[GRAPHIC] [TIFF OMITTED] 97999.045
[GRAPHIC] [TIFF OMITTED] 97999.046
[GRAPHIC] [TIFF OMITTED] 97999.047
[GRAPHIC] [TIFF OMITTED] 97999.048
[GRAPHIC] [TIFF OMITTED] 97999.049
[GRAPHIC] [TIFF OMITTED] 97999.050
[GRAPHIC] [TIFF OMITTED] 97999.051
[GRAPHIC] [TIFF OMITTED] 97999.052
[GRAPHIC] [TIFF OMITTED] 97999.053
Mr. Putnam. Thank you very much. You all are really opening
up some fascinating pieces for us to explore on questioning.
Our next witness is Dr. William L. Scherlis. Dr. Scherlis
is a professor in the school of computer science at Carnegie
Mellon and a member of CMU's International Software Research
Institute. He is the founding director of CMU's Ph.D. program
in software engineering. He is principle investigator of the 5-
year high dependability computing project with NASA in which
CMU leads the collaboration with five universities to help NASA
address long-term software dependability.
His research relates to software assurance, software
evolution, and technology to support software teams. He first
joined the CMU faculty in 1980, after completing a Ph.D. in
computer science at Stanford University and an A.B. at Harvard.
He interrupted his career at CMU to serve at DARPA for 6 years,
departing in 1993, as senior executive responsible for
coordination of software research. While at DARPA, he had
responsibility for research in strategy and computer security,
high performance computing, information infrastructure, and
other topics that we would be shot if we disclosed. He has
served as program chair for a number of technical conferences
including the ACM Foundations of Software Engineering
Symposium, and he has more than 70 scientific publications.
Welcome to the subcommittee. You are recognized.
Dr. Scherlis. Thank you very much. Mr. Chairman and
members, I appreciate the opportunity to appear today to
discuss R&D for information technology. I'm going to make the
case to you that strategic Federal IT R&D is now more important
than before, and that we need proactive leadership to move it
forward. We rely on IT systems pervasively in our economy for
national security, for health care, and for the operations and
safety of our infrastructure. The industry and research
community have made rapid progress in the capability,
performance, and interconnection of IT systems. But despite
this rapid progress, software and IT generally remain immature
as engineering disciplines. We continue to struggle with
quality challenges related to cyber security and software
dependability. We do not yet know how to achieve high levels of
quality in critical systems without huge sacrifices in
capability and flexibility and huge costs to test and inspect.
For both cyber security and software dependability we are
not in a good state. In cyber security, our stop gaps of
firewalls, spam filters, intrusion detection and the like are
not slowing the growth in exploits and vulnerabilities. This is
a chart from the CERT that indicates the number of incidents
that have been reported year over year. We are not succeeding
in evaluation and validation. The Common Criteria ISO 15408,
for example, does not yield guarantees regarding an absence of
malicious code. In software dependability, we cannot in general
make strong promises on the basis of testing and inspection.
The coverage is not good enough. We supplement this by looking
at how the code was developed and who did it. But these are
poor proxies. We cannot, in general, fully evaluate software
artifacts directly. Even when we can see every line of code, we
cannot make promises about the systems we build.
It is tempting to conclude that this bad state is intrinsic
to IT; that things are the way they will be--for example, that
because we get e-mail, we will also get huge volumes of spam.
Or, that we are at a plateau--the pace of innovation is slowing
down, the 1990's are over. Or, that the sheer mass of the
deployed base will inhibit any fundamental change--can we
switch the entire country over to drive on the left-hand side
of the road?
These conclusions are counter to the historical truth of IT
for the past 40 years. There has been a constant technology
revolution under the hood in operating systems, data bases,
client server architectures, networking, languages, and so on.
The research community, the successful IT companies and their
customers all know how to handle this pace of change because
they have been doing it for so long. In many areas, it is
happening right now. But not in the most critical areas related
to quality. We are almost complacent with our extreme
vulnerability.
However, there is reason to hope for the future. There are
promising research results in the pipeline that bear on these
major challenges. For example, more secure network protocols
and services. Improved identity and authorization management.
Techniques for the direct evaluation of software. Securable
architectures for resilient designs.
Given this, it's tempting to think that with the large R&D
budgets the IT industry will take care of this and the
government can step back. And this is wrong. Part of that
historical truth of the past 40 years is that the Federal
Government has consistently been an active player and leader in
that process. And I'm going to give you four reasons why, and
these reasons have to do with why industry does not in general
look beyond more than one or two product cycles out.
First, many of the most significant research results that
bear on IT quality are nonappropriable. That means that their
value diffuses rapidly across the market. It cannot be
retained, it becomes a public good. Only government is going to
sponsor this work. Bill Gates, for example, talks about a tool
that is now used to reduce the frequency of blue screens. This
tool is based on technologies that were developed a decade ago
by my university colleagues and sponsored by NSF and DARPA:
binary decision diagrams and model checking.
Second, the early definition of standards has a
particularly significant role in IT. This is the so-called
prenormative work most vividly illustrated by the role of the
IETF in the early days of the Internet and the role of the W3C
more recently. The world of E-commerce is held together by
standards such as TCP, IP, XML, HTTP, and so on.
Third, government is a major IT consumer. It needs to
collaborate with its entire simply chain, just like the auto
industry. Long ago, DARPA exerted profound influence on
networking and operating systems and processor design to create
an amazingly scalable foundation for network centric warfare
and modern command and control generally. It worked directly
with the vendors, the innovators, and the researchers
throughout the DOD supply chain.
And, four, the main input to the IT food chain is
university research and education. Without the people and
expertise and the innovative attitude, we have nothing.
There is another reason. IT innovation leadership is
pivotal to the future of our country. I'm here from Pittsburgh.
We can argue about the strategic necessity of leadership in
steel or in consumer electronics, but IT innovation leadership
is different. We cannot give it up. It's a driver of
productivity, as Alan Greenspan has noted. It is a principal
force multiplier in defense. And, perhaps most importantly, we
still see no bounds on the potential for creating new value,
new kinds of capability and cognitive powers. The frontier of
innovation will continue to exist well beyond the frontier of
commoditization. It will be our future for a long time.
My conclusion is that we need proactive Federal R&D
leadership. We need both basic science and mission motivated
Federal R&D in order to retain our leadership position and to
address the new challenges that we face. In the public private
partnerships--the collaborations of industry, academia, and
government--the government must be a full partner. I appreciate
the opportunity to appear today.
Mr. Putnam. Thank you very much. I appreciate it, Dr.
Scherlis.
[The prepared statement of Dr. Scherlis follows:]
[GRAPHIC] [TIFF OMITTED] 97999.054
[GRAPHIC] [TIFF OMITTED] 97999.055
[GRAPHIC] [TIFF OMITTED] 97999.056
[GRAPHIC] [TIFF OMITTED] 97999.057
[GRAPHIC] [TIFF OMITTED] 97999.058
[GRAPHIC] [TIFF OMITTED] 97999.059
[GRAPHIC] [TIFF OMITTED] 97999.060
[GRAPHIC] [TIFF OMITTED] 97999.061
[GRAPHIC] [TIFF OMITTED] 97999.062
Mr. Putnam. Dr. Squires, you are going to have a few
moments to collect your thoughts to defend HP's honor with
regard to your R&D budgeting. We are going to recess for a
moment while we go have one vote. It should be a fairly brief
recess, and we will return shortly. So everybody sit tight,
enjoy your orange juice. And the subcommittee will stand in
recess.
[Recess.]
Mr. Putnam. The subcommittee will reconvene. I apologize
for the delay.
Our final witness for this panel is Dr. Stephen Squires.
Dr. Squires is the chief science officer and vice president at
Hewlett-Packard. He is also a Special Government Employee
Expert Consultant for the Department of Defense through the
Defense Advanced Research Projects Agency [DARPA]. In that role
he is a member of the Intelligence Science Board and also
served on the Defense Science Task Force on Defense Roles and
Missions for Homeland Security and other special working
groups. Previously, he worked for NSA for 15 years on IT
systems.
We look forward to your testimony. You are recognized, Dr.
Squires.
Dr. Squires. Thank you for inviting me to testify. I
consider it an honor and a privilege to be here to discuss
these critical issues.
I want to focus on one main issue which was in the letter
that invited me, and that issue is how the investments serve to
protect this Nation and position the United States as a leader
in the information technology arena.
My answer to this critical question is based upon my
understanding of the history of IT, my own direct experience
and expertise in the most advanced research and development
application programs focused on the most challenging problems
facing the Nation, and my own vision of the future.
The best way I can think of starting is to reference a
paper titled ``As We May Think'' by Vannevar Bush in July 1945.
It was written in his role as Director of the Office of
Scientific Research and Development, coordinating activities of
some 6,000 leading American scientists in the application of
science to warfare. You should actually take a look at this
article. It presents an extraordinary vision filled with all
kinds of interesting examples, including one example called the
``memex,'' which is essentially the Internet with a web of
linked objects.
The history of information technology is dominated by
fundamental devices from the invention of the transistor, the
integrated circuit, and the microprocessor, along with many
other devices for the past 50 years and an extraordinary
collection of systems developed through multiple layers of
modules, structures, and massive amounts of software to support
a wide range of applications.
Information technology has become increasingly pervasive.
It is hard to imagine life without it. Our national defense,
homeland security, depend on it, in addition to our critical
infrastructure, the economy, and the future of science and
technology.
IT industry and its applications have become a
multitrillion dollar sector of the global economy that is
recognized as enabling a new global dynamic. Information
technology as generally viewed today appears to be a commodity,
but it is not. The larger IT companies claim to have
multibillion dollar R&D programs, which they do. But it is also
very important to understand the operating point and the time
horizon of those programs.
It is natural for many people familiar with normal
technologies to believe that there has been more than enough
U.S. investment in the future of information technology and
enough is enough. Let me say now in the strongest possible
terms that I believe that such a belief is fundamentally
misguided and in my opinion dangerous.
The entire field has been through multiple revolutions and
extraordinary advances that have been made across a wide range
of science and technology areas. But with all of these advances
and all of this investment in the past 50 years, we are really
only at the beginning of a much longer process.
As the limits of what have become conventional integrated
circuits are reached, new technologies with its revolutionary
implications are emerging at the atomic scale in the form of
nano-technologies. New nano-devices can be integrated into new
kinds of things such as new kinds of nano-integrated circuits
with extraordinary properties, and properties that go beyond
just computing to include new kinds of storage, sensors,
effectors, and new ways to act with the physical world,
including biological.
The advances in these new devices will enable new kinds of
modules, new kinds of units of replication, and present new
challenges, challenges which simply will not be overcome by
conventional industrial R&D. These, the new kinds of systems
which we can imagine happening and emerging over the next 50
years, or even the next 10 or 20 years, will be far more
dramatic than anything we have seen in the last 50.
Let me just give you some examples of the role of
information technology that go beyond normal market trends: The
role of information technology in things like critical
infrastructure of the country, in science and technology
itself, in national defense, homeland security, trusted
information sharing, protecting individual privacy, and the
most important of all, protecting the future of civilization.
Let me briefly sketch four alternative futures. I'm just
going to call them red, orange, yellow, green in the context of
an idealistic vision, blue.
Red is essentially pre-Internet technology.
Orange is essentially an attempt to extend the red to cope
with the emerging Internet revolution.
Yellow is essentially an attempt to apply commercial
Internet technology to the challenges of the Internet.
And green is essentially the development of fundamentally
more advanced technology than commercial Internet technology
for the purpose of achieving strategic advantage. Such systems
have more advanced cybersecurity than the commercial Internet.
And then all of this is set in the context of blue, which
is an idealistic vision of the future which I call ``intrinsic
trust'' and that is the essential distinguishing characteristic
of the fundamental advance needed for the future of information
systems themselves.
The most challenging problems provide insight needed to
establish the most effective advanced research agendas. The
ideal of blue is essential to guide the advanced research
agenda for green.
Given its own market forces, the information technology
system will simply be stuck on the yellow brick road. The
insights needed to create effective advanced research agendas
emerge from interdisciplinary interaction among science,
business, homeland security and national defense. The
interactions are more critical because the need for public-
private systems to interoperate over a wide range of modes are
all dependent upon critical and pervasive interoperable
information systems capable of trusted information sharing
while protecting privacy.
I believe it is essential that the U.S. Government continue
to invest in advanced research and information technology
focused on protecting this Nation and ensuring that the United
States continues to be the world leader in information
technology. The future leadership depends upon continuing
advances in science and technology at a time when information
technology itself is not only becoming critical and pervasive
but itself going through its own reinvention process.
Mr. Putnam. Does that conclude your remarks?
Dr. Squires. Yes.
Mr. Putnam. Thank you very much, and I apologize for making
you wait.
[The prepared statement of Dr. Squires follows:]
[GRAPHIC] [TIFF OMITTED] 97999.063
[GRAPHIC] [TIFF OMITTED] 97999.064
[GRAPHIC] [TIFF OMITTED] 97999.065
[GRAPHIC] [TIFF OMITTED] 97999.066
[GRAPHIC] [TIFF OMITTED] 97999.067
[GRAPHIC] [TIFF OMITTED] 97999.068
[GRAPHIC] [TIFF OMITTED] 97999.069
[GRAPHIC] [TIFF OMITTED] 97999.070
[GRAPHIC] [TIFF OMITTED] 97999.071
[GRAPHIC] [TIFF OMITTED] 97999.072
[GRAPHIC] [TIFF OMITTED] 97999.073
[GRAPHIC] [TIFF OMITTED] 97999.074
Mr. Putnam. We will begin with the questions; and I will
begin with you, Dr. Squires.
Given that there will always be a finite amount of money
available for Federal research and development, are we
prioritizing the use of that funding in the most efficient
manner?
Dr. Squires. I don't think efficiency is the way to think
of it. If you try to optimize the systems so that there is
minimum duplication and maximum efficiency, you are very likely
to lose the most interesting and innovative system that the
world will see. It is all right for the innovation process of
the national R&D agenda to be slightly loose and informal. You
always need the flexibility for a bright young mind to come up
with a new idea without having to go through a long-drawn-out
proposal process. Start soon.
The other thing is I think that there's a general structure
that I've found that is useful in trying to set research
agendas and that is to focus on the fundamental technology
trends, understand their fundamental limits, and when you begin
to identify a limit, look as hard as you can for effective
alternatives. If you take a look at every major advance in
information technology, you discover that it is a result of
that process.
Mr. Putnam. As someone who transcends the public and
private sector, is there adequate collaboration between the
two?
Dr. Squires. There is a lot of collaboration, but I don't
think it's adequate, and I will give you the main barriers. I
think that the mechanisms in place for setting agendas and for
holding full and open competitions and for negotiating the
actual agreements takes too long. It is way too burdensome and
actually tends to be a disincentive to people and individuals
and universities and companies, my own personal opinion,
working in that way.
The phrase that I sometimes use in discussions like this,
and I say our adversaries do not have the advantage of our
procurement system.
Mr. Putnam. Dr. Fossum, do small startups and smaller
universities with less well-established research relationships
with the Federal Government, startup companies that have a
great idea born in a garage, do they receive adequate
attention, adequate opportunity to compete with the big, well-
established companies that are out there?
Dr. Fossum. In theory, yes. Is it a level playing field on
paper? Yes. But there's an awful lot of networking,
connections, and experience that go into knowing how you
actually successfully get a grant proposal through and how you
actually go through a procurement and become a participating
competitor, so to speak. I think it is a very large threshold
for some to get across.
Actually, when you mention universities, there are over
1,800 4-year accredited colleges and universities and
professional schools in this country, and only 80 of them
receive 71 percent of the Federal funds for R&D. It shows you
the concentration of these funds. It is an enormously
concentrated world.
Mr. Putnam. Dr. Lazowska, is that a concern, the
concentration?
Dr. Lazowska. I think equity and distribution and
participation is always a concern. But I think the numbers that
were just cited need to be interpreted in light of the fact
that many of those 1,800 4-year schools do not purport to have
a research program. They are purely educational institutions.
So we need to make sure that research funds are equitably
distributed to those best positioned to carry out the work. And
I think Federal agencies do a reasonable job of that. There are
also programs that ensure adequate research funding to States
that don't have perhaps a full complement of research
institutions.
In terms of prioritization, I would just remark that the
PITAC committee in 1999 did a fairly thorough analysis and
concluded that the Federal investment in IT R&D, compared to
other fields, was dangerously inadequate; and they proposed a
ramp-up which has not been nearly met. In fact, as Mr. Clay
pointed out, the NITRD funding will actually decrease in the
proposed budget. So I think many Members of Congress are coming
to the conclusion that the Federal R&D portfolio has become
unbalanced. Perhaps that doesn't mean we're investing too much
in some areas, but it means we're investing too little in
others, and you have identified a number of them today.
Mr. Putnam. The imbalance being toward the biological
science and CDC, health?
Dr. Lazowska. I would say the imbalance is against
information technology. If you look at the role IT plays in
national security, in advancing the sciences, in driving our
economy, we are investing a relatively tiny amount of Federal
money in creating the next generation of advances.
Mr. Putnam. Dr. Scherlis, do you have anything would you
like to add going to this?
Dr. Scherlis. I guess I would.
Your first question to Dr. Squires was concerning
prioritization within the R&D portfolio, and I think it's worth
clarifying a little bit from the outside my understanding of
the NITRD process. It is not a top-down process. It involves a
combination of vision and mission. The mission agencies who are
collaborators within the NITRD process identify their needs and
priorities on an agency mission basis. This is combined with
input from the research community, who drive the process on the
basis of their invention and imagination and desire to explore.
It's the juxtaposition of those two things that really creates
the innovative magic that several of my colleagues have spoken
of earlier in this hearing.
Mr. Putnam. Is the allocation of R&D skewed too heavily to
defense and defense-related research? Dr. Scherlis?
Dr. Scherlis. Actually, it's interesting, the history of IT
in this country has a history that's really largely been driven
by a combination of Defense and the National Science
Foundation. It is a combination of those two organizations. And
it has been interesting growing up in IT to see that IT
researchers and scholars have adopted defense metaphors. They
speak in terms of survivability and command and control. DOD
has had a profound influence; and, as a consequence of that,
the mission needs of defense have been very well met through
the research community.
This has worked well because of the very farsighted, broad
attitude of the DOD basic science investors, those who invest
6-1 and 6-2 funds in the R&D world. It's the supply chain
management story. They work throughout the supply chain, not
just with the prime contractors and the systems integrators but
with the vendors and the innovators and the inventors who feed
that supply chain. They do it in a way that provides value well
beyond the defense mission. There is a leverage in that story.
The leverage is that DOD is able to buy off-the-shelf
components and systems that they can apply directly in their
mission. We have pervasive applications, spreadsheets, word
processors, operating systems; and these are increasingly
incorporated as components into systems. They are pervasive not
just in our offices and homes but also in our national
infrastructure and critical national security systems. So what
has happened is that the pervasive systems have become
critical. This is an inevitable and positive outcome. It's the
nature of IT that we have come to this point. It makes these
problems, in some ways, much more significant and challenging.
Mr. Putnam. Dr. Fossum.
Dr. Fossum. I think one of the questions that you raise is
about what is defense R&D. We don't really know how to divide
R&D. R&D doesn't know a home agency or a discipline. It can go
across all kinds of areas. Let me give you an example.
I happened to be at the Army 1 day talking to a gentleman
there whom I literally asked ``What is your major technology
challenge?'' And he said they needed a ``miniature, long-life,
anticorrosive fuel cell'' to put in every piece of equipment
they deployed in the field. And we put in the term RaDiUS--
``fuel cell'' into RaDiUS and got hits in six or seven
different agencies including NIH. We shortly were looking at an
actual description of ``miniature, long-life, anticorrosive
fuel cell'' research at NIH for the artificial heart. That is
why it is very hard to talk in terms of ``health'' research
versus ``defense'' research. R&D doesn't know boundaries like
that. Discovery doesn't recognize those boundaries.
Mr. Putnam. Dr. Squires.
Dr. Squires. One thing I'd like to add to what Dr. Scherlis
mentioned is the fact that throughout the history of advanced
research in science and technology leading to advanced products
there's a pattern that having people able to understand and
focused on the hardest problems of the time with the resources
and flexibility to solve those problems is what leads to great
invention. You don't make great advances by looking at the easy
problem. You don't make great advances by doing what everybody
else is doing. You make great advances by going beyond what you
normally can think of doing and trying to invest the future.
It turns out that because of the nature of the American
system, the American economy, the role of defense in the United
States and around the world, the defense and national security
system, the homeland security system of the country has among
the most challenging problems for information technology and
science. It's just a fact, and that is a tremendous source of
insight and motivation, and the people who make the investments
and set up the research agendas normally do it in such a way
that the technologies are as much as possible dual use. Because
it doesn't do you much good to have an advanced technology if
you can't afford it. So it's a very subtle, complicated and
important relationship between the public sector, the private
sector, the civil agencies in the government, the National
Security Agencies, and the government how that works in the
community.
Mr. Putnam. Clearly, it is an important role for defense to
play in research; and it's allowed us to be on the cutting
edge. Frankly, it is something to be very proud of. But it's
just interesting to think about the breadth of research that
occurs in the name of defense, whether it is an MRE in food
preservation or it is training a dolphin to go seek out a mine
and everything in between that leads us to things like the
Internet, things like GPS that are now in every brand-new
suburban sold and all of these other things.
Dr. Scherlis.
Dr. Scherlis. One of the reasons why it is important to
focus on mission R&D is that the needs of Federal agencies
often anticipate the market in terms of their demands for
capability and quality. In areas where they follow the
marketplace they should generally follow the marketplace with
respect to acquisition as well. But in IT, the history has
always been that many mission agencies, not just the DOD but
the Department of Energy, NASA, other agencies, have needs
that, frankly, go beyond the needs that are evident in the
marketplace at any given moment. When they invest in a dual-use
fashion, they get this tremendous impact from the investment
because it creates an economic stimulus as well as a response
to the agency needs. That actually pays off for the mission
agencies. It's an important and sometimes essential payoff for
those agencies.
Dr. Lazowska. I'd just add one other fact to this, and that
is the track record over many decades is that it takes about 15
years from the invention of an idea to when it's exploited in a
billion dollar industry. OK? So what that means in some sense
there is no such thing as just-in-time research if you are a
commercial enterprise. Companies cannot afford to be investing
in innovation that is not going to pay off for 10 or 15 years.
I'm a shareholder, and you are a shareholder, and that is not
the way we make investment decisions.
That speaks to the role of the Federal Government and the
Federal agencies because the mission of the Federal agencies
require these advanced technologies. They support the
innovation which in many times makes its way into the private
sector, but it is many years later, and that is why the Federal
Government has such an important role.
Dr. Scherlis. Forgive me for prolonging the discussion, but
I have to add one more point. Mission agencies enjoy another
advantage, which is that they can afford to be farsighted. They
don't have to make a quarter-by-quarter ROI case for every
research investment they make. They can anticipate their needs
through a planning process, and they can respond to these needs
through their R&D mechanisms. And that, combined with the fact
that they don't need to explicitly appropriate the value that
they create through that R&D investment, creates a tremendous
synergy that allows them to be much more aggressive and to get
more leverage for their funding.
If you look at the level of DARPA funding over these many
years as compared with R&D funding in any one of the major IT
companies, it is relatively low. But DOD gets enormous impact
for that investment because they are investing in a leveraged
way. They are applying this supply chain management trick of
investing where they see they can get the maximum impact for
that investment in the long run.
Mr. Putnam. Are we still the cutting-edge Nation for basic
research?
Dr. Lazowska. There have been many claims in recent years
that we are losing that edge, and the concern of PITAC has been
that we are losing that edge in information technology because
the level of investment has not kept pace with the
opportunities of the field and the increasing demands on the
field. So I think there is reason for concern, but it's very
difficult to measure.
For example, there was something in the papers a few months
ago talking about the number of physics publications in Europe
and Japan versus North America. I think we have to expect that
the rest of the world is going to start contributing at a level
comparable to what we're contributing. I urge to you think
about the areas in which this Nation cannot afford not to be
the world leader, and I would assert that information
technology is one of those areas. We cannot afford not to be
the world leader in information technology, because it drives
everything else. It drives every other field of science, every
other field of engineering. It drives the economy. It drives
defense and security. We can't afford to fail to be the world
leader in this one field.
Mr. Putnam. Give me another field that we can't afford not
to be the world leader in.
Dr. Lazowska. You will have to convene another panel,
sorry.
Mr. Putnam. If I had CDC, would they say we couldn't afford
not to be the world leader in gnomics and biotechnology and----
Dr. Lazowska. I firmly believe that 30 years ago I stumbled
into the field that underpins all other fields, OK? So I firmly
believe that this field is No. 1 in terms of the leverage that
it offers.
Dr. Scherlis is talking about leverage, and this is a field
that offers enormous leverage in all other fields. You can't do
advances in the biomedical sciences these days or in health
care delivery without advances in information technology. You
know, astronomy is digital imaging and data mining of the
images.
Mr. Putnam. You are not supposed to say data mining
anymore. Ask these DARPA guys.
Dr. Squires.
Dr. Squires. Don't ask me that question.
I think next on my list is, obviously, nano-technology,
because of its fundamental implications. But simply investing
in nano-technology without the context of its transforming
effect on all of information technology and all other science
would be a mistake. As I said in my testimony, as extraordinary
as the advances have been in the last 50 years, they were
enabled by a relatively small number of fundamental device
inventions and a massive number of systems structures and a
massive number of software technologies; and what we have
today, as wonderful as it is from my sort of perspective of the
future, is a really very small scale prototype of what it could
really be 50 years from now.
Mr. Putnam. In the green world or the blue world?
Dr. Squires. Yes, the blue curve. My blue sky vision says
we need to be in this world of the green, and we need to get
off the yellow brick road. The rest of the world has seen what
the U.S. process of invention and innovation has done. We need
to get on the new curve. Otherwise, what is the risk of having
some other part of the global community decide to get on the
green curve?
Mr. Putnam. It is a tremendous risk. But I mean, in 1985,
we were all being told to go on to speak Japanese and that the
Japanese were buying up the whole Nation and they owned the
motion picture studios. And because they owned the motion
picture studios, it was the end of America as we knew it.
We go through these periods, and I am not in any way
arguing with a distinguished group like yourself that it is not
important for us to continue to be the world leader in IT. I am
just trying to play devil's advocate here.
Dr. Scherlis. I made a point in my testimony about the
necessity of IT innovation and leadership as compared with
other engineering disciplines. The issue is where, globally, is
the focus of innovation and how important is it to have that
focus of innovation. The reason that IT is interesting, as
compared with other building materials--we can think of
software as a kind of building material--is that other building
materials can only scale up so much. You can only build a
building that is so tall before various laws of physics start
to impede our ability to build it taller. With software, we
don't see any such natural limits.
The Windows operating system I believe is now 50 million
lines of code. Who contemplated 50 million lines of code even
20 years ago? Impossible. And there is no reason why we can't
go from 50 million to 50 billion and to create systems of
tremendous cognitive power, for example, that can translate
languages or be autonomous robots or cars that can drive
themselves. There are many such visions.
My point is that these are no physical limits in the world
of IT that impede us from addressing those aggressive visions
directly. And in fact, as Steve Squires just said, let's focus
on the hardest problems. What's interesting, in fact, is that
partly is what DARPA is doing right now. They are focusing on
the hard problems of cognition and how to build smart systems
that can learn. That is a very good topic.
But we also need to be focusing on the bread and butter
issues of how can we make promises about the systems that we
build, how can we make dependable systems and secure systems?
These are hard problems, and these are the impediments to
scaling up.
Mr. Putnam. You said earlier there are absolutely no bounds
for innovation to IT.
Dr. Scherlis. No physical bounds. There are only
intellectual bounds.
Mr. Putnam. Does anyone disagree with that? Anyone wish to
add to that?
So how do you jump off the yellow brick road, Dr. Squires?
Is it, as Dr. Freeman said, education, turning out more
engineering and computer scientists and graduate degrees that
are home grown? What is the trick to maintaining our leadership
role in IT?
Dr. Squires. I think it is important to have an effective
framework for thinking about these future worlds that go beyond
just the nano-devices and beyond just the applications so that
you are actually able to effectively organize the different
disciplines to work with each other.
My favorite one, which I wrote a little bit about in the
testimony, has five major layers from the bottom up: devices,
modules, structures, virtualizations, and applications. Each of
these is a major discipline in itself which works with all the
other disciplines; and the most important thing is to look in,
pick your favorite framework--that's my favorite one--and try
to understand what the fundamental trends and limits are.
So, for example, in devices, the fundamental trends and
limits at the device level, these are actually running out of
the ability to build integrated circuit technology with
increasing performance and cost-effectiveness as we do it
today. It costs billions of dollars to build the next VLSI
product. You can save hundreds--many orders of magnitude when
we transition to nano-scale self-organizing technologies. That
may be 5 years before we get the first devices, but we
certainly need the first new devices in that area.
Assuming that you can do that, you have to think what are
the new modules, units of replication, which, if we could have
that new manufacturing capability, would we choose to have? If
you had that, then what would be the new virtualizations, what
would be the new system structures, and what would be the new
applications?
The interesting thing is what has happened is the
transistor was invented and all the kinds of wonderful things
happened after that. What we learned is that having multiple
layers of the system work in parallel is way better than having
them work in series. So if you have the basic and applied
sciences working across the full range, from devices to
applications through those intermediate levels, working on
always the most important problems so the feedback would be not
just produce the papers, produce real stuff, real system
prototypes, real prototype products, real products which early
adopters can use sooner rather than later and get the feedback
to the system, you have the potential to bring the future into
reality sooner rather than later.
So I view this whole investment strategy as a kind of time
machine. What you are actually doing with Federal R&D
investment is getting an earlier view of the future than
anybody else can, Getting it in the minds of the best
scientists and engineers and businesspeople in the country and
making it available to the United States and all the people of
earth sooner rather than later. And doing one more thing:
Through the American system, providing the incentives so it is
used for good, as opposed to used for something else. So I
can't imagine life without being on the frontier.
Mr. Putnam. Dr. Lazowska.
Dr. Lazowska. I think the simple answer to your question is
``support IT R&D by Federal agencies at the level that the
President's Information Technology Advisory Committee
recommended in 1999.'' What you see from this graph that I
showed you earlier is that we're increasing Federal R&D
overall. What you see from this graph is that Federal R&D in
information technology has flat-lined. It has fallen far below
the PITAC recommendation, and it has flat-lined, and the good
news is these are very small numbers. We are talking about only
hundreds of millions of dollars a year. That is not an
inconsequential amount of money, but on the scale in which the
Federal Government operates or the Federal R&D operates, it is
inconsequential.
Our government has failed to prioritize this field. Simple
as that. What you understand very well is the role that this
field plays in our economy and in all other fields.
Mr. Putnam. Dr. Fossum, anything to add?
Dr. Fossum. Just sort of a random thought that came by. I
look back at the history of some of the things that the U.S.
Government has done in R&D over the last 50 or 60 years, and
what it took to do it. We have a history of when something is a
true priority nationally of essentially having a major program
or an incubator on it. Go back and start with the Manhattan
project and look at the form they took. If IT R&D is very
important to this Nation, which I think we all agree it is,
maybe we need to look at some models that we used before,
rather than rely on the current funding streams of the current
agencies. Maybe we need to rethink how better to pull the parts
together.
Mr. Putnam. You raised a good point, and I have forgotten
most of what I learned in junior high and high school science.
All of you operate, as Dr. Squires put it, on the frontier.
There is a great deal of apprehension and concern about our
inability to attract young people into the math, science, and
engineering fields; and yet as far back as the Manhattan
project we were pretty well co-opting the world's talent
anyway. We were offering them freedom, a safe place to live,
work, raise their family, and have a future and apply their
brilliance to productive, hopefully peaceful things, although
we could certainly have a whole other hearing on that.
But how great a crisis is that? How big a threat is that?
And how much attention should we be paying to it? Is it a
natural demographic occurrence that China will produce an
annual increase of more engineering degrees than the sum of all
of our schools? Or is it truly a crisis in American higher
education? We'll start with Professor Lazowska.
Dr. Lazowska. One thing I would say we should make sure
that our immigration policies are in line with the sorts of
goals you have articulated. That is, do we allow the best
students from around the world to come to the United States and
get training? And do we allow them to remain in the United
States once trained? If they return to their home countries, do
we use them as agents of international cooperation or perhaps
are we closing our borders and preventing these smartest minds
from around the world from coming to us and learning?
Mr. Putnam. Anyone else before we lose power?
Dr. Scherlis. I just want to say that it is a serious issue
and that many universities are struggling to develop strategies
to attract the very best students into these fields. In our
programs, we continue to get the very best students, but, in
many other programs, there are challenges. We find, for
example, that applications from overseas have gone down
considerably because of this friction at the border.
So, yes, this is definitely an issue. The most fundamental
element of our supply chain is the people who populate it, and
I think we need to take it up explicitly.
Mr. Putnam. Dr. Squires, you and Dr. Scherlis, y'all were
both recruited very early in your careers into government
service and research?
Dr. Squires. I was, essentially as a freshman
undergraduate.
Dr. Scherlis. I was on the faculty at Carnegie Mellon when
I was recruited.
Dr. Squires. But I recruited him.
Dr. Scherlis. I thought it would be an easy desk job for a
couple of years, and I could write papers in my spare time. But
seriously, it was the most exciting and demanding thing I ever
did. I stayed much longer than the usual tenure, and I continue
to strongly recommend service. You talked about this with the
earlier panel. We all feel the sense of possibility and
opportunity to really do something significant for the Nation.
There are many good reasons to take this up.
Mr. Putnam. So high turnover is a fact of life in these
fields? I mean, you said you stayed longer than normal.
Dr. Scherlis. The IPA law, Interagency Personnel Act,
allows a maximum stay of 4 years for somebody rotating in from
a university or a State or local government. In fact, what I
did was to stay on the IPA for 4 years; and then I rejoined as
a senior executive government employee.
Typically, rotators from universities to NSF or DARPA or
other agencies will stay between 2 and 4 years and then return
so that they don't lose continuity in their home institutions.
Dr. Lazowska. I think it helps to be part of the research
community, to work with the research community. So both NSF and
DARPA have had great success with recruiting top members of the
research community into being office directors and program
managers for a period of time and then sending them back.
Mr. Putnam. Dr. Fossum, and then we're going to wrap it up.
Dr. Fossum. DARPA and NSF are stellar examples of this.
That's not the case as I can see with a place like NASA where
IT is very critical and where they have a large budget for R&D,
but they are not real good at R&D partnering because they tend
to deal more with contract-driven R&D, then grant-driven R&D.
And the world you are talking about is where universities can
coordinate with the Federal Government to cooperate and
leverage the talent at universities is a ``grant driven'' world
for the most part. In that world, you are talking about the
``science program'' at Department of Energy. You are talking
about NSF. You are talking about NIH.
So just like we have a problem with the substantive
border--perhaps we also need to take a look at how the R&D
dollars move, and also, where the laboratories in the Federal
Government will open their doors to various and sundry people.
DARPA is world famous for this. World class. Maybe we need to
use them as an example to teach a few other parts of the
Federal Government how they might do that, too.
Mr. Putnam. You pushed a button.
Dr. Scherlis.
Dr. Scherlis. Yes. I just want to present an alternative
perspective about collaboration with NASA.
Dr. Fossum. Oh, they do some. No doubt.
Dr. Scherlis. I lead a project with NASA that involves
Carnegie Mellon and five other universities, MIT, University of
Southern California, University of Washington, University of
Wisconsin, and University of Maryland. And that project is
structured as a cooperative agreement which allows us to
collaborate directly with NASA mission managers, mission
engineers, and intramural researchers. We find that to be a
very successful structure for collaboration. And I also want to
note that at DARPA, at least to my knowledge, most of the
relationships that they build with researchers at universities
are framed as contracts or cooperative agreements. The nature
of the vehicle through which the collaboration is undertaken,
is I think, less important than the culture and horizon of the
sponsoring organization.
Mr. Putnam. Dr. Fossum's total agreement is noted for the
record.
Dr. Fossum. Yes. I just wanted to make one point. And the
only point I was trying to make is that if you look at the
proportion of, for instance, cooperative agreements, which are
the vehicle that should be used, in some agencies, they haven't
learned how to use them quite to the extent they might want to.
NASA is an example of such an agency.
Mr. Putnam. One of the things that I talk about in my
Rotary Club speeches is when you look at the success of the
American military and exponentially ahead of our competitors in
a variety of fields, it's really because of the investments
that the American people have made with their hard-earned tax
dollars for decades that yields tremendous military prowess
that then translates into the commercial sector. And, you know,
people don't normally realize it until they hear it that they
are shareholders in the success that ultimately not only raises
our living standards but saves lives.
Our society is not particularly good at recognizing and
rewarding smart people, and yet it's the brilliant people in
laboratories, in universities and in the Federal Government
that just do tremendous things to make our lives easier,
better, healthier, more productive and worthwhile. And one of
the side effects, the positive side effects among many in the
dot-com boom was that it kind of made it OK to be smart again.
Working hard and being smart and attentive and paying attention
to the sciences and math and computers, would get you a billion
dollars or more in the case of some of them. And, hopefully we
can find some way to tap into that generation of young people
who have grown up seeing that and encourage them to continue to
pursue their studies and academics and make it OK to be smart
again.
It's been a pleasure having such a smart panel showing
their wisdom with us.
Before we adjourn, I want to just convey to you the
subcommittee's deepest appreciation for your accommodating us
and dealing with the voting schedule. Your testimony is very
important to our better understanding of Federal R&D, and we
appreciate you.
In the event that there may be additional questions that we
do not have time for today, the record shall remain open for 2
weeks for submitted questions and answers.
Thank you all very much. The subcommittee is adjourned.
[Whereupon, at 4:38 p.m., the subcommittee was adjourned.]
[Additional information submitted for the hearing record
follows:]
[GRAPHIC] [TIFF OMITTED] 97999.075
[GRAPHIC] [TIFF OMITTED] 97999.076
[GRAPHIC] [TIFF OMITTED] 97999.077
[GRAPHIC] [TIFF OMITTED] 97999.078
[GRAPHIC] [TIFF OMITTED] 97999.079
[GRAPHIC] [TIFF OMITTED] 97999.080
[GRAPHIC] [TIFF OMITTED] 97999.081
[GRAPHIC] [TIFF OMITTED] 97999.082
[GRAPHIC] [TIFF OMITTED] 97999.083
[GRAPHIC] [TIFF OMITTED] 97999.084
[GRAPHIC] [TIFF OMITTED] 97999.085
[GRAPHIC] [TIFF OMITTED] 97999.086
<all>