Firefight Over The Weapons LabsJohn Carey and Robert D. Hof
At first glance, DYNA3D is a rousing success story of beating swords into plowshares at Uncle Sam's vast weapons laboratories. Developed by the bomb wizards at Lawrence Livermore National Laboratory in Livermore, Calif., DYNA is sophisticated computer software for simulating the workings of nuclear warheads. Because it can be used to study how metal deforms, it's now also being applied to designing everything from stronger beer cans to safer cars. And lab researchers have spun off Livermore Software Technology Corp. to create and sell commercial versions of the software. "DYNA is a classic case where the national labs have really helped American industry with a technology that was developed for weapons," says laboratory engineer Robert G. Whirley.
Yet there's a darker side, critics say. Lab scientists are increasingly working directly with corporations that use the software to improve it. The lab's software code is free. As a result, charges Bence Gerber, marketing vice-president of Livermore Software, the national lab is acting not as a friend of U.S. industry but as a formidable rival. "It's very difficult to compete with a product that's free," Gerber complains.
This spat over DYNA3D is part of an enormous dilemma: what to do with the national laboratories. Over the past half-century, the U.S. government has created hundreds of facilities for research into nuclear weapons, science, medicine, and agriculture. Ranging from the Army's Cold Regions Research Lab in New Hampshire to Agriculture's Cotton Ginning Research Unit in Mississippi, these have a combined budget of $22 billion a year, 30% of the $72 billion federal R&D budget. Now, in both government and industry, there's "a growing feeling that we're not getting an optimal return on our investment," says President Bush's science adviser, Yale University physicist D. Allan Bromley. The quandary is most acute for the huge nuclear weapons labs--Livermore, Los Alamos, and Sandia (table). "When we needed the muscle to win the cold war, the ideas that made that possible came out of these laboratories," said President Bill Clinton during a May 17 visit to Los Alamos, N.M.
Today, however, the weapons labs are scrambling to find new missions. Each is more than twice the size of the research operations at giants such as AT&T Bell Laboratories and IBM, or at major universities such as Stanford. Bristling with supercomputers and other advanced technology, the weapons labs consume about $3.57 billion per year of the Energy Dept. budget--more than the National Science Foundation spends on basic research. But while they designed and built the most advanced nuclear weapons, "what's still to be proven is their ability to do anything else," says the Alfred P. Sloan Foundation's Hirsh Cohen, who chaired a high-powered panel that studied the labs in 1987. These facilities, adds University of California Berkeley chemist Bruce M. Novak, are "dinosaurs struggling to become mammals."
The fate of the three labs depends on whether they can make that transition. Many in Washington and lab officials themselves want to use them to boost U.S. competitiveness. Livermore Associate Director Roger W. Werne imagines federal scientists helping companies develop clean cars, powerful chips, and new environmental technologies. "We believe in our hearts and souls this is important for the country," says Andrew B. White Jr., head of Los Alamos' renowned supercomputing unit. So does the White House, which has made links between industry and these laboratories central to its R&D policy. Their technology can "revolutionize the economy," says Clinton.
LENGTHY MAZE. This new role raises troubling issues, however. "The labs have enormous resources, but they're not that well connected to our civilian industrial economy," says Clinton's science adviser, John H. Gibbons. Indeed, companies must get through a time-consuming bureaucratic maze to hammer out joint research agreements with the labs. Moreover, corporate executives and even venture capitalists say they rarely find useful technology on the shelf.
Most damaging is the argument that these labs will have little bottom-line impact on U.S. industry. "We see a lot of utility in working with the labs, but we are kidding ourselves if we say that will make Corning much more competitive," says David A. Duke, vice-chairman and R&D chief at Corning Inc. Some executives, such James A. Ionson, research director at Polaroid Corp., believe the billions of dollars could be better spent shrinking the deficit, reducing corporate taxes, or improving the workforce through better science and engineering education. If Washington truly wants to transfer lab technology, adds Joseph G. Morone, director of Rensselaer Polytechnic Institute's Center for Science & Technology Policy, "it would be better simply to give money to companies to hire lab scientists."
Proponents say the labs haven't had a fair chance to prove their worth. Although Congress passed laws in 1986 and 1989 directing them to push technology to companies, "there was a lot of resistance to tech transfer under Reagan and Bush," says Senator Jeff Bingaman (D-N.M.). Now that Clinton's team has made it clear that the labs' future size depends on how valuable they are to industry, "these folks are getting religion real fast," says Dan E. Arvizu, head of technology transfer at Sandia. The three weapons labs now have 192 cooperative R&D agreements with industry, up from only 15 in 1991. "This is a wonderful way to bring a huge amount of expertise to us in a way we can afford," says Richard L. Kegg, vice-president for technology at Cincinnati Milacron Inc., which has an $8 million program with Lawrence Livermore to develop more accurate machine tools.
Success stories may remain rare, however, critics contend. Making research pay off with new products is hard enough within a company. And the labs have a starkly different culture--a tradition of secrecy and performance at any cost. "It's tough for our guys," concedes Kay V. Adams, director of technology transfer at Los Alamos. "How to do it cheaper and faster was not the mentality of the labs."
This huge gulf between industry and federal research explains why billions of taxpayer dollars spent at the labs in the 1970s on alternative energy sources produced few practical results. It also explains why the innovative technologies they've concocted--such as ultra-lightweight aerogels--are commercial wallflowers, too expensive to produce. And why companies such as Quadrax Corp., a Portsmouth (R.I.) high-tech composite maker, haven't been helped. "A few years ago, Livermore did some analysis of our materials," says CEO Richard A. Fisher. But the security-conscious lab kept the results secret--even from Quadrax. "It didn't do us much good," Fisher says.
THORNY QUESTION. Many companies can't afford to contend with the labyrinthine bureaucracy. "The labs' knowledge could be of considerable help to us," says Jim Simon, engineering vice-president at machine-tool maker Giddings & Lewis Inc. in Fond du Lac, Wis. "But we don't have the administrative layer for dealing with the government--nor do we want to add it."
Then there's the thorny question of how much these labs can offer. After massive taxpayer investments, "there jolly well ought to be something," says J. David Roessner, professor of public policy at Georgia Institute of Technology. One measure of the commercial promise of their technology is the number of spin-off companies. With only 40 or 50 each, the big labs fall far short of major universities such as Stanford or Massachusetts Institute of Technology, which have spawned hundreds.
Nor are these labs attracting many venture capitalists. The reason, speculates Don Valentine, general partner of Sequoia Capital in Menlo Park, Calif., is that what really matters in starting companies is entrepreneurial zeal, not great technology. "I'd rather have second-rate science and people with phenomenal passion than the reverse," he says. "And in my experience, there are few people like that in the labs." Lab officials aren't surprised. "The labs have been a closed institution for a long time," says Livermore's Werne. "It has not been our charter to spin things off."
When it comes to technology, moreover, "there are a limited number of nuggets out there," says Georgia Tech's Roessner. They include supercomputer simulations, advanced materials, precise machining, and high-powered lasers. As industry mines these nuggets, "this will level out soon," he predicts. More troubling, a survey he did for the Industrial Research Institute shows that companies looking for technology and information tap other U.S. companies, universities, foreign companies, and private data bases before turning to the labs.
Perhaps most difficult is the issue of competition. DuPont Co. is upset that the Pentagon's Wright-Patterson lab in Dayton makes a high-tech polymer virtually identical to one it sells. And companies making high-powered semiconductor lasers don't like a growing laser effort at Livermore, which they say uses taxpayer subsidies to win government contracts and commercial work. "The labs are killing off the industrial base by taking the funds that would otherwise be available to companies like McDonnell Douglas and Spectra Diode," fumes a McDonnell Douglas Corp. executive. Richard W. Solarz, Livermore's program leader for advanced lasers, replies that his team only wins when it offers clearly superior technology.
Complaints about stealing business or favoring companies could derail the labs' civilian effort. "We see a real mine field out there," says Al Narath, president of Sandia National Laboratories. As more companies get involved, "there are going to be winners and losers--and the losers could sue the labs," says a congressional aide. "It could be dicey."
PROOF WANTED. If that's not enough, Congress is considering major changes. George E. Brown Jr. (D-Calif.), chairman of the House Science, Space & Technology Committee, has introduced a bill that could consolidate the weapons labs and convert at least one to civilian work, such as environmental technologies. Energy and Pentagon officials counter that all three labs are needed to maintain the nuclear stockpile.
What's clear is that a day of reckoning is looming. A new Congressional Office of Technology Assessment report suggests that the labs must prove they offer something better than universities, nonprofit labs, or companies. "We are shifting from an entitled existence to one based on performance," says Sandia's Narath. At the very least, this means some downsizing or consolidation is "inevitable," said Energy Chief Hazel R. O'Leary at a recent hearing.
How much the labs will shrink depends on how useful they prove to be. Not everyone is hopeful. "The joke around here is that eventually Lawrence Livermore will have a few thousand bureaucrats overseeing the work of two scientists," says Livermore astrophysicist Hugh E. DeWitt. If that happens, these dinosaurs will become extinct.