ADVERTISEMENTS
Learn about diversity at ASEE
ASEE would like to acknowledge the generous support of our premier corporate partners.

 FEATURES

By Art Pine
+ Photo by Michael Temchine
OPEN TO WILD IDEAS - DARPA Director Regina Dugan confers with Marine Brig. Gen. Robert Hedelund, vice chief of naval research, before a House Armed Services subcommittee hearing on March 23, 2010. Photo by Michael Temchine
COVER STORY

OPEN TO WILD IDEAS

An engineer brings ‘blue sky’ research back to DARPA.


Say you’ve had a brainstorm you think might lead to a totally new kind of weapon or to a networking breakthrough with national security potential. What better place to seek financing than the Defense Advanced Research Projects Agency, crucible of the Internet and Global Positioning System? Until recently, if you were a university-based engineer or scientist, getting DARPA support would have been difficult, if not impossible. For much of the past decade, the agency has turned to industry contractors it thought would produce results more quickly. Academics complained that the old DARPA, a risk-taking champion of pure research, was gone.

But today, DARPA is back in the market for the best and the brightest university researchers. Its director, Regina E. Dugan, a mechanical engineer and former university researcher, has been visiting campuses, reconnecting with old contacts, recruiting young people, and re-establishing ties with academe. More fundamentally, she is starting to reconfigure the agency to make it friendlier to postdocs and faculty members, while at the same time opening up avenues of research bound to excite younger scientists.

“It’s an attempt, in a sense, to go back to DARPA’s golden age—the idea that the agency can sponsor totally outside-the-box research and see what kind of interesting stuff results,” says Peter Harsha, director of government affairs for the Washington-based Computing Research Association, whose members are primarily academics.

That think-outside-the-box mentality — along with the reliance on researchers from universities, defense corporations, and the four armed services — has been a DARPA hallmark since its inception. And it has been one of the big factors in DARPA’s success, both scientists and military officers agree.

Created almost on a whim in 1958 as part of the Eisenhower administration’s frantic response to the Soviet Union’s surprise launching of Sputnik, the Advanced Research Projects Agency (as it was known at birth) was designed to oversee the nation’s then disparate missile and space efforts. But it soon was pushed aside by the newly formed National Aeronautics and Space Administration, which gobbled up the space mission and has kept it ever since. Left without a portfolio — and hobbled by a slow start — ARPA turned to research on potentially cutting-edge technology that was too esoteric for other agencies to touch.

Over the next 50 years, DARPA — which added the word “Defense” to its name in the early 1960s — gained fame for coming up with an array of spectacular innovations that have revolutionized the way the military conducts wars and have led to breathtaking changes for civilian society as well. DARPA oversaw the research that led to ARPANET, the forerunner of the Internet. DARPA projects led to the Global Positioning System, which uses a constellation of satellites to help locate aircraft, ships, and vehicles and to make it easy for them to navigate.

The agency spearheaded the effort to develop stealth technology, which enables aircraft and ships to evade enemy radar. DARPA also propelled the nation into delving into serious seismography, shepherding the creation of a worldwide network of satellites and sensors eventually used to help detect hidden nuclear testing. A 1984 DARPA project eventually morphed into today’s Predator and Global Hawk unmanned aerial vehicles. These aircraft now number in the hundreds and have become a major weapon against insurgents in Afghanistan and Iraq. More recently, DARPA coordinated a series of projects that led to the development of a revolutionary prosthetic hand that enables war fighters who lost a limb in combat to regain almost all the functions of a real hand. For decades, amputees had to settle for hooklike pincers that gave them only limited capability.

A Start-up Culture

In truth, DARPA doesn’t actually conduct the research on any of its projects. Its cadre of program managers, based at its headquarters in Arlington, Va., plucks ideas from applications sent in by university and industry researchers. DARPA then sets the parameters for the projects and finances them. In so doing, it nurtures technology that neither industry nor academia could develop without government financing. “When all is said and done, most of what they’re asking for is typically stuff for which there’s not a pot of gold at the end,” says Dean Kamen, the New Hampshire-based entrepreneur-inventor.

“It’s clear that DARPA is back in the game.” — Peter Harsha, director of government affairs for the Washington-based Computing Research Association. - Part surveillance satellite, part aircraft, Vulture is intended to stay aloft for five years.

Crucial to the agency’s success is the DARPA culture, which traditionally has encouraged researchers to explore long-shot ideas at will without fear of being tagged as failures if they don’t come up with anything useful. Almost as often, one thing leads to another, and the project yields valuable results that no one had envisioned. Although the agency is part of the Defense Department, it’s relatively unencumbered by bureaucratic requirements, and its projects needn’t be linked directly to weapons or military needs. The agency’s lifestyle is relaxed. At work, employees can wear anything they like, including T-shirts, jeans, and sneakers.

Peter Lee, who headed the computer science department at Carnegie Mellon University before a yearlong stint at DARPA, describes “a really fun environment — something like a start-up,” but one that’s intensely challenging. “The amount of intellectual pressure we’re put under all day, every day, is significant, and beyond anything in my professional experience,” Lee told the New York Times in an interview last April. He recently left the agency to join Microsoft.

Most important to some DARPA enthusiasts is that, unlike a typical government bureaucracy, the agency imposes term limits on its people. Program managers typically stay no more than three years before returning to academia or to industry — a revolving-door approach designed to ensure that the people overseeing projects will be up to date and receptive to new ideas.

Program managers who arrive from universities encounter an entirely different set of demands from those placed on an academic. “As a faculty member at a university, you’re expected to be deeply knowledgeable about a narrow set of technical areas, but as a DARPA program manager, you have to have a sufficient depth of knowledge in a very broad area — how those technologies could come together to provide a military capability,” says Donald Leo, associate dean for research and graduate studies at Virginia Tech. From 2005 to 2007, he headed chemical and biological defense research at DARPA. “It was an exciting job,” he says.

The sometimes zany-sounding projects that DARPA takes on — combined with a strong sense of secrecy intended partly to protect the military advantage that comes from surprise — have inspired a mythic reputation. Author Michael Belfiore dubbed the agency “The Department of Mad Scientists” in his book of that name published in 2009. The dust jacket describes it as “America’s greatest idea factory” — a “juxtaposition of outlandishly ambitious goals with basic, commonsense business principles and everyday problem solving.”

Academic Estrangement

DARPA’s estrangement from universities began after Anthony J. Tether took over as director in 2001, under the George W. Bush administration. He moved to shift the agency’s workload away from university-based researchers — and from long-term, open-ended projects as well — toward research by defense contractors that promised quicker results. Under Tether, DARPA adopted an array of new guidelines that, while not explicitly barring university-based scientists from taking a big part in DARPA projects, made it significantly harder for them to do so.

The agency shortened the period for which it would provide financing; classified almost all research from its inception; set an annual “go, no-go” review for all projects, which made it tougher for universities to sign graduate students to four-year contracts; and barred noncitizens from some projects — a constraint for many university science departments, where typically half of the graduate students are foreign born. It also insisted on the right to review articles about such research before they were sent to scholarly journals, even when the subject matter and data were not classified. “That was antithetical to most universities,” Harsha says. Many schools prohibit their researchers from submitting articles to sponsors of research prior to publication.

As a result, opportunities for academic researchers became severely constricted, Harsha says. While DARPA’s overall budget grew, its financing of university research fell by half, after adjustment for inflation, between 2001 and 2008. Deprived of agency funding, the Massachusetts Institute of Technology and other universities cut back their computer science departments. “A lot of people had to go to industry,” says Anant Agarwal, a professor of electrical engineering and computer science at MIT.

“More important, the character of DARPA’s funding changed as well,” Harsha says. “It used to be that university scientists could direct their own research and determine how it plays out. But under the new model, DARPA would fund large defense firms as general contractors for such research, and university researchers would be subcontractors.”

That situation limited what university researchers could do, he says. “It meant that there was a large body of brilliant minds at universities who no longer were working on military-related technology.” Many feared that DARPA would lose its ability to launch the kind of longer-term research that had produced the Internet and stealth technology.

Ten red balloons were launched in 2009 in a contest exploring social networks as information-gathering tools.

Reopening Doors

Dugan, the first woman to lead the agency, has begun to relax some of these strictures. She herself previously served as a DARPA program manager from 1996 to 2000, toward the end of the pre-Tether era. With degrees from Virginia Tech and Caltech, she’s no slouch in the wild ideas department. Among efforts that led to her being named program manager of the year in 1999 was the “Dog’s Nose,” a field-portable system for detecting the explosive content of land mines. Becoming director in July 2009, she renewed the agency’s focus on the kind of “blue sky” basic research that had earlier given DARPA some of its eye-popping achievements. She tapped Peter Lee to head a Transformational Convergence Technology Office to focus on social networks, synthetic biology, and machine intelligence — attractive fields for younger engineers and scientists.

Dugan also lifted the security classification from many of DARPA’s basic research projects and revoked the requirement for prepublication review of articles on unclassified research, an approach Defense Secretary Robert M. Gates has urged the rest of the Defense Department to follow.

Most important, Dugan gave program managers at headquarters broader authority to set the rules for each project, relax restrictions they think aren’t needed, and clarify regulations that inadvertently create problems for researchers. And program managers now have authority to bring in non-U.S. citizens on many projects. While restrictions still apply to some projects, the overall result has been to open more doors to academics, says Kaigham Gabriel, DARPA’s deputy director.

In computer science, to name one area, the change is noticeable. “It’s clear that DARPA is back in the game and is funding a wide range of computer science projects,” Harsha says.

Dugan’s new approach to the treatment of basic research has already made a difference to university-based researchers, says Robert M. Berdahl, president of the Association of American Universities. “From the standpoint of universities interested in basic research, her leadership has been gratifying and, I think, very effective,” Berdahl says. Yet DARPA’s struggle to balance the involvement of university researchers, scientists from industry, and experts from the military services isn’t over. Dugan herself says the agency wants to keep all three of its constituencies working together.

Linda P.B. Katehi, chancellor of the University of California at Davis and an electrical engineer, cautions that restoring the once close working relationship between DARPA and the universities will continue to be a work in progress. “DARPA is making some very positive changes, but it will take time,” she says.

Universities can still expect to face stiff competition. Kamen’s DEKA Research and Development Corp., for instance, has a spate of DARPA projects on its platter: a robotic arm to complement the cortex-controlled hand; small hydrofoils to help Navy SEALs swim longer distances; and gadgets that can catapult soldiers to four-story heights.

One of Lee’s efforts was to reach out to a broad range of possible candidates — “unusual circles of innovators,” as he told Defense Systems — in part by making DARPA’s proposal requests more succinct and welcoming, with fewer technical specifics.The agency already maintains an open-source channel to let large numbers of outside designers submit proposals for software. Now it’s creating a similar channel for designers of vehicles, aircraft, and spacecraft.

Current DARPA Projects


A new helicopter threat-alert system. It tracks the shock wave from incoming small-arms fire and returns fire with pinpoint accuracy from a U.S. chopper.

A surveillance and imaging system that gives ground troops a continuous wide-area, high-resolution color photo of the entire theater, combined with computer tracking of all major movements by the enemy.

A new long-range hypersonic missile that researchers say will be launched with booster rockets and then glide 3,000 nautical miles to deliver bombs and other weapons.

A technique for producing antivirus vaccines — for protecting against biological attack — in just 30 days, instead of the three to six months required previously. The new procedure involves inserting a portion of the genetic sequence of the virus into special tobacco plants. Until now, vaccines have been grown inside eggs, and take 180 days.

New thin sheets of graphene made from unrolled carbon nanotubes (one-billionth of a meter thick) promise to break through the limits imposed by traditional silicon microelectronic devices, producing equipment that has far higher operating frequencies and requires far less power. That means radar systems with 10 to 15 times the current range.

A search to unlock the mysteries of the magnetic field sensing systems that enable birds to navigate long distances and give dogs their keen sense of smell.

Plans to speed innovation in application software, or apps, to enable war fighters to use low-cost hand-held devices for special-purpose computing.

A photonic microchip that uses light instead of electrons to transmit data, enabling speeds and efficiency levels similar to fiber-optic links.

A rocket that can send a 500-pound unmanned aerial vehicle loaded with intelligence and surveillance instruments to any point on the globe within one hour and then fly slowly around the areas and beam back intelligence for up to seven hours.

A study of sled-dog metabolism changes during the Iditarod Trail Sled Dog Race in Alaska that keeps the dogs’ energy high.

Research into why a breed of Asian snakes can glide long distances in the air and actually turn.

New Research for a New Era

One of the agency’s boldest new forays is its effort to grasp the potential of social networking both to threaten national security and to expand DARPA’s technological capability. Gabriel calls it a rapidly emerging “opportunity area.” In a widely followed contest in 2009, DARPA publicly invited thousands of university scientists to use social networking to locate 10 red balloons that it distributed around the continental United States. The experiment was designed to help understand how information is disseminated through social networks. The winner was a team from the Massachusetts Institute of Technology, which located all 10 in just eight hours and 56 minutes.

But that’s only a small part of the agency’s interest in social networking and crowd-sourcing. Dugan and Gabriel were intrigued by “Trapster,” a system by which motorists can subscribe to speed-trap alerts on a BlackBerry or other hand-held device; and by North Korea Uncovered, in which some 35,000 Internet users identified military bases and economic infrastructure shown on a Google Earth map of the secretive nation.

With the United States engaged in two wars that present immediate battlefield needs, DARPA recognizes that “blue sky” research can’t be the sum total of its mission. Dugan and Gabriel, both of whom founded and headed technology manufacturing companies, are intent on shortening the time required to turn innovations that spring from research into usable technology or hardware. Rather than taking 10 years to develop a nearly perfect ground combat vehicle, for example, DARPA’s new manufacturing model aims to put it in the field in two years and then improve it in increments as necessary. “We’re being more responsive [and] adaptive” that way, Gabriel says.

In a change that presents opportunities for engineers, DARPA plans to spend $1 billion over five years streamlining the manufacturing process to eliminate production delays, surprises, and cost overruns. It seeks to enable other industries to emulate the semiconductor industry, where, Dugan says, the progress from design to manufacturing is seamless.

Finally, DARPA has streamlined many of its own management practices, paring the mountain of unobligated funds that the agency has accrued each of the past few years. The backlog had drawn criticism from congressional appropriations committees, which complained that too often DARPA asks for funds and later ends up not using them.

In an era of tight budgets, a federal research agency that wants to spend money faster can only be good news for university engineers and scientists.

 

Art Pine is a Washington-based freelance writer and former Pentagon correspondent.

 



TOPˆ

 


ASEE
© Copyright 2011
American Society for Engineering Education
1818 N Street, N.W., Suite 600
Washington, DC 20036-2479
Web: www.asee.org
Telephone: (202) 331-3500