2. "Uncle Sam Wants You!"
Actually, I didn't join the Army, but I did take a break from my academic
career to join the Advanced Research Projects Agency (ARPA) of the Department
of Defense. This is an organization unique within the Federal government
consisting of some of the hardest working technical staff, supported with
the biggest budgets for research and development. This tiny agency of less
than 150 Program Managers had an FY95 Presidential Budget of approximately
$2.5 Billion. And I have never worked as hard or for as long hours as I
did during my two years at ARPA-and that includes graduate school and life
as an untenured faculty member.
ARPA was established in the wake of the Sputnik "disaster" for
American science and engineering. Its mission has since been to insure that
America avoids future technological surprise by her military-and increasingly
economic-adversaries. It is little known that ARPA ran the early space program,
before the establishment of NASA. ARPA oversees extensive research programs
in remote sensing, stealth technology, and smart weapons. However, at least
in our academic circles, ARPA is best know for its long term investments
in the underlying electronic and computer systems technologies. These have
seeded the development of major industries in packet switched communications
networks, the Internet, time sharing, computer graphics, workstations, UNIX,
computer-aided design software, high performance computing, and the list
continues. These developments were justified to the Pentagon brass by their
contribution to ARPA's defense missions: world dominating command and control
systems for the military.
ARPA's mission was well articulated by Dr. Gary Denman, APRA's Director
until earlier this year:
- Pursue imaginative and innovative research and development projects
having significant potential for both military and commercial applications
and, in some cases, projects with solely military applications.
- Support and stimulate a national technology base that serves both civilian
and military purposes through technology sharing. By serving both dual-use
and military purposes, we hope to increase productivity of both sectors.
- Manage and direct the conduct of basic and applied research and development
that exploits scientific breakthroughs and demonstrates the feasibility
of revolutionary approaches for future military applications. This will
include advanced technology having future civilian applications.
- Stimulate increased emphasis on technology demonstrations in defense
systems and subsystems.
This statement reflects the Clinton-Gore administration's priorities with
respect to the conversion of defense industries. There is a strong emphasis
on so-called "dual use" technology development: technologies with
application in both the commercial and defense sectors. The current Congress
has different priorities, and there has been a re-emphasis on technologies
that meet military needs on the one hand, and which address basic science
on the other. Fortunately, the Joint Chiefs have an understanding of the
potential advantages of advanced information technologies for supporting
the America's warfighting capabilities, ingrained by the successes in the
Gulf War. It remains to be seen how this will affect ARPA's research mission
in the near future. [5].
The keys to ARPA's success are its unique non-bureaucratic culture and its
commitment to long-term projects. Its organizational culture grants technically
oriented program managers substantial latitude in formulating their research
and development programs.[6]. Occasional failure is expected-it
is a maxim at ARPA that if some projects do not fail, then the agency is
not investing in enough high risk ideas. The program managers play a unique
role as brokers between Defense customers for technology and the research
community that is developing the new technologies. To keep the ideas and
enthusiasm of the program managers fresh, the agency has always been characterized
by relatively high turnover: an ARPA assignment is a "tour," not
a career. For me, two years immersed in this intensive environment were
more than enough.
One of my more memorable ARPA experiences was the briefing I gave to then
Deputy Secretary of Defense John Deutsch. He requested a review of several
big ticket ARPA programs, including High Performance Computing and Communications
(HPCC), as a prelude to a budget cutting drill. I felt that I had twenty
minutes to lose a $100 million of core computer system research funding!
Fortunately, the presentation went well, and the only tough question I received
was what happened to Thinking Machines. Another colleague at ARPA was not
so lucky. After showing Secretary Deutsch a particular example of whiz-bang
technology, his response was "Am I supposed to be impressed by this?"
This is not the kind of response you want to get in a high stakes briefing
like this!
The second key to the agency's success is its provision of substantial support
over extended periods of time in order to achieve truly significant results.
The focus is on teams developing proof-of-concept and early prototype systems.
Much of the research is DoD budget category 6.2, so-called "applied
research." It is not uncommon for ARPA take on projects that cost tens
of millions of dollars (or more). The first stealth aircraft prototype is
one notable example of a breakthrough technology.[7].
This stands in stark contrast to the 6.1 "basic research" pursued
by such DoD organizations as the Office of Naval Research (ONR), the Army
Research Office (ARO), and the Air Force Scientific Research Office (AFSRO).
The latter fund mostly small single investigator grants in the same spirit
as National Science Foundation (NSF). And they have suffered significant
cuts by the budget ax.
ARPA is world-renowned for its research programs in information technology.
These have stimulated most of the advances in computing and communications
technology over the last several decades. A common question at ARPA, called
by some the "twenty year criterion," was phrased as "what
will be the technology developments that we will be proud of when people
look back twenty years from now." ARPA has left its mark on scalable
computing systems, intelligent systems, networking technologies (wireless,
optical), and advanced software development. Many of these will provide
the critical underlying ubiquitous computing capabilities for the National
Information Infrastructure (NII): telecommunications, databases, advanced
computer systems for information access in the 21st century.
So it should come as no surprise that the Clinton-Gore Administration would
turn to ARPA to help it get "wired" into the National Information
Infrastructure. And that is where the adventure really begins.
Footnotes
5. A major reorganization of ARPA has recently taken place.
See http://www.arpa.mil for the latest details.
6. It offered the perfect playground for someone like
me, interested in understanding long term technical directions.
7. We once received mail from the ARPA Comptroller asking
whether anyone knew about a Boeing aircraft, as its invoice had just come
in. There were several e-mail responses along the lines of "That's
stealth technology for you-did you lose the airplane in the parking lot?"
[Prev] [Top]
[Next]
Randy H. Katz, randy@cs.Berkeley.edu, Last Updated: 21 December
95