The Advanced Research Projects Agency-Energy -- better known as ARPA-E -- has one of the federal government's most intimidating lobbies.
Twenty-four volumes of the "Encyclopedia of Chemicals Technology" stare down from one corner, while well-worn books with titles like "Dimensional Analysis" and "Biothermal Fluid Sciences" are stacked in various piles on adjacent shelves. A visitor without a doctorate might be limited to one option: the Dalai Lama's "The Art of Happiness."
All belong to ARPA-E Director Arun Majumdar, a former academic who has led the fledgling research agency through its first $400 million in research grants.
The books are a remnant of Majumdar's days as a professor and researcher at the University of California, Berkeley. But they also exemplify the academic atmosphere at ARPA-E, which funds high-risk, high-reward projects with a goal of creating a new energy industry.
The agency has become a testing ground for the kinds of organizational changes Energy Secretary Steven Chu would like to spread throughout DOE. In some ways, Majumdar, 48, is Chu's kindred spirit, espousing similar theories on hiring practices and management goals.
"We are both technical people. Science and engineering is sort of in our blood," Majumdar said. "So we would have some, I would say, similarity of thinking about issues."
In a recent interview at ARPA-E's headquarters, Majumdar described the hiring policy he thinks has led to the agency's success. With the help of laws that exempt ARPA-E from some federal hiring requirements, Majumdar can handpick his team. Scientists, businessmen and engineers swap the usual drawn-out federal hiring process for temporary positions that last about three years.
"They're motivated to do really well because their future career depends on what they do out here. They know that they cannot stay," Majumdar said. "So there's a clock running. There's a clock ticking going on if they have to do it in a certain number of years. And they have to do well because if they do not, they may not get hired."
In Majumdar's view, that power should be extended to other DOE programs -- an opinion that is supported by the President's Council of Advisors on Science and Technology. In a report last year, the council praised ARPA-E for demonstrating "that it is possible to have an agile and effective technology program within DOE."
"It has demonstrated that streamlined processes can attract superb individuals and teams to make proposals," the report found.
"To accelerate the transformation of the energy system, excellent people need to be engaged in government programs and work on energy technologies. To sustain this success, it will be important to extend this streamlined approach to project execution as well."
Chu has also expressed a desire for more flexibility in DOE's hiring, pointing out that ARPA-E employees are sometimes described as "higher quality than the people getting the funding."
In February, Chu made Majumdar the acting undersecretary for energy, giving him oversight of some of DOE's most prominent programs and perhaps hinting at Chu's plans for the wider agency.
But it is still unclear whether Majumdar will take the position permanently; DOE has been silent on the issue, and Majumdar will only say that he leaves the decision to Chu and President Obama.
In the meantime, Majumdar has taken on both jobs with enthusiasm, joking that he sometimes sleeps four or five hours a night. He also manages to fit in trips every few weeks to Berkeley, where his wife and two teenage daughters still live.
That kind of cheerful dedication suited him well at the Lawrence Berkeley National Laboratory, where he was the director for the Environmental Energy and Technologies Division for about three years.
Ashok Gadgil, who was Majumdar's deputy and now heads the division, said Majumdar could often be seen talking with employees next to a whiteboard filled with formulas.
"He truly threw himself into this job," Gadgil said. "He committed himself deeply to the division and he spent an enormous amount of intelligence and effort in making sure he could address the most urgent problems."
Majumdar was not originally drawn to a management job, preferring to stick to research. But he said Chu -- who was the director of the Berkeley lab at the time -- "twisted" his arm to lead the division.
Gadgil remembers Majumdar preparing for his presentation to the hiring board by talking to every employee in the division about potential challenges. The board hired him almost immediately.
"I wasn't sure I really wanted to do that or not, but I took the plunge," Majumdar said. "It's a lot of fun. Management is, you know, there is a human aspect to it, there's an organizational aspect to it. There's frankly a vision and an action plan to get to your vision."
Gadgil said Majumdar has the right temperament for ARPA-E, where part of the job is expecting the majority of research projects to fail.
"Besides the intellectual vitality and enormous intelligence and knowledge that he brings to the job, he also has very, very good interpersonal skills," Gadgil said.
"And that's another really important key aspect in being successful in shaping an organization early on so it is not only focused on getting results but also energizing people he's working with."
But Majumdar might soon be limited in what he can accomplish at ARPA-E.
Though Democrats and Republicans have praised the agency, GOP lawmakers have also targeted it for deep cuts. The vast majority of the agency's funding so far has come in the form of stimulus funds.
The House has proposed $50 million for the agency for the rest of the year -- a significant cut from Obama's $300 million request. Even the Democrat-controlled Senate has floated a scaled-down figure of $200 million. The White House's $550 million proposal for fiscal 2012 seems politically impossible.
"It is certainly a concern of ours. Money does matter at some point," Majumdar said, but he declined to detail how the agency would survive such deep cuts.
"Frankly, we are absolutely delighted to be in the budget and a lot of people worked very hard -- members of Congress as well as staffers -- to keep us in the budget."
But proving ARPA-E's success can be tricky, and Majumdar admits that results might not be seen for 10 or 20 years. But he pointed to the recent announcement that six projects had attracted about $100 million in private capital -- about four times as much as ARPA-E put in. The agency has also sparked the formation of several startup companies.
"Now that's the right kind of leveraging of federal taxpayer dollars," he said. "Those are good signs I would say. Not a home run yet, but good signs.