Michael Belfiore
The Department of Mad Scientists
I found this juxtaposition of outlandishly ambitious goals with basic, commonsense business principles and everyday problem-solving inspiring.
Note: D mcandler
This had to involve two major leaps on DARPA’s part, it seemed to me: (1) the flexibility to see that difficult problems might be solved by not just unconventional means but also unconventional people, and (2) the insight to sort out the truly viable approaches from the mere crackpot schemes. What exactly was this DARPA, I wondered,
Actually, that statement comes surprisingly close to summing up much of DARPA’s work: find an area of technology that could go a long way toward serving the needs of the country if improved but that wasn’t getting much attention in the private sector, put some well-considered research and development money into it to get it on its feet, and then cut it loose.
Note: nice
doors were firmly shut to outsiders, this
The community was to include garage tinkerers in addition to the more traditional defense contractors. “I think that was the key insight that he had—that that would be of tremendous value,” said Whitaker.
The trick to the Challenges was to find a way to do business that satisfied the requirements for the government’s contracting system but that nevertheless included those garage inventors and other entrepreneurial types who ordinarily would never have had a chance to work with the government. Offering prize money instead of simply funding participants outright was the key to making it work—that and managing the motley crew of inventors, university researchers, and military contractors on their way to the starting line.
The point is that innovators in a given field of endeavor see what is possible, and are inspired to push the envelope further themselves. “We always had this idea, you see, in order to expand the field, what we needed to do was create the visual of robots actually operating safely in traffic, and preferably with human drivers as well,” Whitaker explained to me. He figured people wouldn’t be that impressed by robots interacting with one another on those streets. But robots interacting successfully and safely with human drivers—now, that was another thing altogether.
The goal was to remove the giggle factor.
More than building a robot that was going to achieve any particular goal, the goal in this program was really the demo. The demonstration of what was possible.”
“We took the job of driving and broke it down into lots and lots of different pieces,” said Whitaker, “and then took those pieces and decided some of them were too hard, or not necessarily too hard, but there were things we weren’t going to test this time.”
What’s more, the event would act as a kind of practical robotics conference and get the best researchers in the field on the same page. After the event, those researchers should be able to collaborate or share information with one another much more easily.
The conflict between the need to clearly define operational requirements while leaving them vague enough for future applications created what Norm Whitaker called “a dynamic tension” between DARPA and the teams. “See,” Whitaker explained to me, “the teams, if they can get you to define the problem well enough, then they’ll build sort of a special-purpose machine that will do just that. If you tell them that every turn is going to be between eighty-five and ninety-five degrees, that’s what you’ll get, exactly that.” To produce autonomous vehicles that would be truly useful to the military, the requirements for the robot race cars had to be less explicitly defined.
“There is no question,” he said to me, “that the challenge introduces a nonlinear rate of accomplishment that is unachievable in a mercenary context or something where you’re paid to play.” There was no better way to push the envelope in any technical field, as far as Whittaker was concerned, than through open, naked competition. “You can’t buy that performance,” he said. “And some of that is because you can’t buy a human heart.”
The race, Tether proclaimed, was a “fantastic accomplishment.” He said that the technologies developed for it were now ready to be developed for production vehicles and that DARPA could now step out of the picture. “DARPA is an interesting organization,” Tether reflected. “We really never finish anything. All we really do is show that it can be done. We take the technical excuse off the table, to the point where other people can no longer say, ‘Hey, this is a very interesting idea, but you know that you can’t do it.’ I think we’re close to that point, that it’s time for this technology to [be furthered] by somebody else.”
The point was there were people actively working on those problems now who hadn’t been before.
DARPA’s greatest asset, Tether told me, is its program managers, and greatly contributing to their success is their limited tenure at the agency. “The term limits are what really make this place,” said Tether. “When someone walks in the door here, there’s a sense of urgency that you don’t find anywhere else in government.”
“A typical DARPA term is three to four years,” program manager Douglas Kirkpatrick explained to me, “and a typical maximum DARPA term is six years.” There’s no danger of forgetting; every program manager has an expiration date printed on his or her badge. “If there was any word I would [use to] describe that is uniform across all DARPA program managers,” said Kirkpatrick, “it’s just how intellectually aggressive we are. It is in part due to the fact that all of us come in seeing the clock ticking…and get the hell out of our way because we’ve got stuff to get done. That’s what we’re about. We’re rather impatient people, and in the government, that’s a good trait.”
“The people who come here have an idea that they just can’t get done where they are,” Tether told me. “They have a feeling that it can be done—not necessarily the knowledge of the exact details…but they have a feeling that it can be done, and if it can be done, that it’s going to be wonderful.”
Incoming program managers “know that they’re not going to get their idea finished in the four years that they’re here,” Tether said. But “they know that they can actually get it started and get it through those first early hurdles.” They know, in other words, that they might be able to make the impossible possible. It’s an irresistible challenge to the right kind of mind.
still comes down to a pitch meeting between the director of DARPA and a program manager who wants to start a pet project. The kind of quick pitch meeting that resulted in the funding of the computer network that became the Internet was the rule, rather than the exception, then and now. With just two layers of management—a program office director and the head of DARPA himself—program mangers soon get used to getting a maximum amount of exciting work done with a minimum of paperwork.
Said Tether later, “The projects that we do are typically those projects [where] the idea could end up with a great capability, but there’s extraordinary little data to prove that the idea could be built…. DARPA’s role and uniqueness is that we will take a bet on that idea where other people will not.”
Back in those good old days, Tether explained, “the program managers always had the DARPA director on the verge of being fired for something.” Back then, people ran around the office “with their hair on fire…thinking crazy-ass things.”
He wanted Tether to return the agency to its roots as a free-spirited home for mavericks and mad dreamers operating on the barely respectable fringes of Pentagon culture, the only place where truly creative thought could thrive.
The idea, as Tether later put it, was to “break down all the stovepipes,” to give government officials the tools they needed to sift through the enormous amount of data already flowing into their offices and detect patterns and flag suspicious activities for human analysts to act upon.
Tags: pink
Tether himself remains unapologetic about DARPA’s focus on applied research. “Even the ARPANET, or Internet as it is known today, had an application in mind and that application directly and indirectly drove the research,” he told me in early 2009. He defends the direction in which he pushed DARPA during his tenure. “Open-ended research is categorized as 6.1 research in DOD and is defined as research without an application in mind.” DARPA’s role during his directorship, he said, was to mine this research for ideas that could be applied.
“DARPA is envied because of the capabilities it has brought to realization far in advance of anyone else,” Tether told me after his departure. “This is due to its focus on bridging the gap between open-ended research—the far side—and the near-side research, which improves what exists today. DARPA program managers hired during my tenure all had the capability when they joined to evaluate open-ended research and determine what new capability was now possible. By the time they left, they learned techniques on how to articulate to potential users why they needed this new capability even if it meant the user giving up current programs and doctrine. This combined ability allowed them to develop and transition technologies that provided new capability for the U.S. and saved many lives. I couldn’t be prouder of them and am honored to have been at DARPA when they were there.”
“We really need to take an incremental flight demonstration approach,” Walker told Tether, “where we can develop some technology, integrate it on a flight-test vehicle, fly it, learn, and incorporate that learning into the next vehicle, fly it, learn, and incorporate that into the next vehicle.”
Like those of his colleagues, Walker’s badge displayed the month and year in which he had to leave the agency.
“Ultimately,” he repeated to me, “the strategic goal is to remove energy as a source of conflict worldwide. What I was most gratified by is that I was allowed to say that publicly. That shouldn’t be lost on people—that DARPA at this point in time is engaged in the strategic energy fight.
Energy, he pointed out in his DARPATech talk, is the single most important factor in waging war. Not just now, but in every conflict in history, he said, “energy has been the limiting factor in all military operations.”
term. In fact, he said he hated it, “because it smacks of focusing inward as opposed to focusing on the world as a whole. When I say energy independence, what I mean is energy independence from petroleum.”
Tags: pink
Seems the program’s goal of converting fully half of the solar energy striking a solar cell into electricity was much, much too ambitious, as far as the attending engineers and scientists were concerned. At the time, even experimental solar cells could barely crack 20 percent efficiency. “We were vilified,” Kirkpatrick told me, “because we were taking the best and brightest from the solar world and setting them off on this absolutely hopeless task, and all we were going to do was to prove that solar was not important.” Far from being recognized as fostering innovations in the field, “we were the enemies of the solar world because we were going to show how useless it was.” The problem, as the other engineers saw it, was that DARPA, through Kirkpatrick, was bent on setting up the best minds in the field for failure, and failure would make the whole field of solar power research look bad. To Kirkpatrick, this was a perfect example of what he saw as the key difference between DARPA and the rest of the science and technology community—namely, academic and industrial researchers.
To put it another way, the researchers were evaluated based on what they published, and not on whether they produced solutions to problems in the real world. “You need, what, four, five, six publications a year to get tenure,” said Kirkpatrick. “You can’t afford to have failures, because you can’t publish failure.” On the other hand, industrial researchers were too focused on return on investment, which made them equally unwilling to take chances on potential breakthrough technologies. Academic and industrial research scientists, to Kirkpatrick’s way of thinking, were thus united by an aversion to failure that hindered their ability to make real breakthroughs.
Nurturing, Kirkpatrick explained to me, was the very antithesis of DARPA’s approach. “DARPA is, ‘Get on the train, suit up, and let’s go play.’ And some folks need a little bit more nurturing before that can happen. You can ask my wife. I’m not the nurturing type.”
the very specific purpose of the program: to reduce the number of batteries that soldiers would have to hump into the field. Extremely efficient solar cells could do that
Tags: pink
When Barnett got into the field, there was no commercial market for solar cells—they had been invented in 1955 for satellites.
Tags: pink
“We’re state employees with no state money by choice,” he told me, “and that makes us extremely entrepreneurial.” In fact, fully 80 percent of the EERC’s funding comes from private companies that hire the group to develop specific, practical technologies. “This place is very Darwinian,” Groenewold said. “Those who can, do. Those who can’t, leave. If you don’t bring in money, you don’t get a paycheck. It’s as simple as that.”
Groenewold, who does a lot of his recruiting from the university, takes a rather unorthodox approach to hiring. He asks each potential hire who makes it through the initial screening process to answer three questions: “What is your dream?” That’s to sort those with fire in their tummy from the oxygen suckers, as Groenewold puts
Note: hire q 2
it. “Do you feel lucky?” Groenewold wants the people who work at the EERC to create their own luck and opportunities rather than wait around for someone to do it for them. And, finally, “How many speeding tickets have you had?” That’s to filter out people who are overly risk averse. “I think a person who has never had a speeding ticket is so risk averse,” Groenewold explained to me, “that they should not work here.”
Note: Hiring
Its mission to equip our nation’s war fighters with a technological edge over their adversaries gives DARPA a razor-sharp focus it would otherwise lack, and its emphasis on quickly moving projects from concept to working prototype while making as efficient use of funds as possible should be a model for research
some of the most useful technological innovations of all time. Key to its success—in addition to its minimal bureaucracy—has been the term limits for its managers and its low overhead.