Meeting New Challenges for U.S. Industry
Sandia's Science Park: A New Concept in Technology Transfer
When Sandia National Laboratories announced plans for a new science and technology (S&T) park, it opened a new avenue for partnership with private industry. For years, the national laboratories have been urged to create better connections with the private sector, the better to make full use of their mission-driven research. They have responded with varied initiatives involving licensing and cooperative research, some successful, others not. Sandia has moved in a new direction, risky to be sure, but with considerable promise of success.
With this venture, Sandia follows the lead of dozens of other institutions, mainly universities, that have created S&T parks in order to improve industry's access to their research. But beyond this common denominator, Sandia's specific objectives are much different from those of a university or a commercial developer. Consequently, the criteria for success are also much different.
Federal labs have been subject to close scrutiny in recent years, most notably from the acerbic 1995 Galvin task force report, which cautioned the labs against staking out new missions at a time when traditional mission areas require a more disciplined focus. Thus, any attempt by a federal laboratory to do things differently raises eyebrows. Is the lab straying from its assigned role? Doesn't this belong with the private sector? Why should a lab care about creating jobs in its local economy? How does this activity contribute to national security? In other words, has Sandia been infected with that dreaded disease known as "mission creep"?
These are legitimate questions. In this era of greater accountability in federal research policy, it is important to specify exactly what the point of a science park is and what the criteria for success should be. My view is that in principle the S&T park does make sense in terms of Sandia's mission. Sandia and the other labs are given money to perform their primary missions. Once we've spent this public money on successful research, it makes sense to share the results with U.S. companies, especially because federal scientists also benefit from interaction with their industrial counterparts. Moreover, the S&T park is a mode of technology transfer that may well improve on existing approaches. What remains to be seen is how the Sandia experiment will work in practice.
The first step in evaluating its success is to develop a clear understanding of what we should expect. Sandia is one of America's largest national laboratories. Its main mission is to ensure that the nuclear weapons stockpile is safe, secure, and reliable. Additionally, Sandia is concerned with protecting the nation against technology-based threats, such as terrorists with nuclear weapons, and with various energy issues. Its annual budget is about $1.2 billion, and it employs 7,600 people, primarily scientists, engineers, and technicians. With facilities in Albuquerque, New Mexico, and Livermore, California, Sandia is operated by a subsidiary of Lockheed Martin under contract with the U.S. Department of Energy (DOE).
Like the other national laboratories, Sandia faces the problem of redefining its mission to match the changed realities of world conflict. No more Soviet Union, hence no more Cold War arms competition. But this traditional threat has been replaced by a plethora of lesser yet serious dangers from terrorists and from hostile countries to which weapons of mass destruction have spread. Do these changes reduce the need for what the national labs have to offer? This is widely assumed but never proven. Conceivably, new world circumstances might even require the expansion of research capabilities.
Nevertheless, budgets are tightening. Whatever the labs' mission, they must accomplish it with less money. That is, they must somehow increase the benefits they produce from a given amount of R&D. This suggests the need to spread these benefits more widely, multiplying the use of each new piece of knowledge and technology. Bringing to society the full benefits of what the laboratories do is, of course, the basis of longstanding legislation that mandates the transfer of lab technologies to the private sector. So far, Cooperative Research and Development Agreements (CRADAs), in which a company or group of companies works together with the lab on a specific project, are the main vehicle for doing this. The Sandia S&T Park is a related attempt to spread and leverage the benefits of mission-related technology.
Sandia's novel approach
Sandia's new concept is straightforward: Set aside nearby land where high-tech startups or new branches of established companies can build research and production facilities that take advantage of proximity to Sandia. This, Sandia expects, will foster the transfer of technology from the labs to the companies. The more technology transfer that occurs, the greater are the benefits emanating from Sandia's laboratories, net of any extra costs of establishing the park. Thus, Sandia research contributes more to the nation. In addition, economic benefits will flow to the local economy, thus generating political support for the lab.
Sandia is applying an established idea in a new way. The institution of the technology park has been around for several decades. University-based research parks were first developed in the late 1950s but proliferated rapidly only in the 1980s. There are about 140 of them in the United States, and about 270 in other countries. Two of the largest and oldest are Research Triangle Park in North Carolina (now with nearly 100 companies) and the Stanford complex in California (originally anchored by Hewlett-Packard).
Sandia is not the only national laboratory to have a science park nearby. There's one in Tennessee, not far from Oak Ridge National Laboratory. In that instance, however, the impetus came from non-lab groups whose primary objective was local economic development, not technology transfer. Consequently, firms there are not necessarily linked with the laboratory, and some are not very high-tech. Los Alamos is planning a research park, though again the main emphasis seems to be on helping the local economy. Although the Sandia park is expected to boost local economic development, its primary goal is to contribute to the lab's national mission.
The 200-acre site near Sandia consists of parcels owned by Albuquerque public schools, DOE, the state of New Mexico, and private owners, all of whom appear pleased with the prospect of seeing their dusty acreage transformed into a high-tech center. Not surprisingly, Albuquerque and the state of New Mexico heartily approve of the prospective development and the consequent spillover benefits of jobs and tax revenues. Overall, however, DOE has not played a major role, and costs to the federal budget are minimal and indirect. All of the participants have their own expectations for the project, but I want to focus on how the project meshes with Sandia's goals.
Economic criteria. For university-based research parks, success is conventionally defined almost entirely in economic terms -numbers of companies started, jobs created, property values enhanced, and so forth. But these criteria are not, or at least should not be, the main focus of Sandia's venture.
For Sandia, success should be measured by how effectively the science park contributes to transferring technology and to accomplishing Sandia's core missions. But before it can achieve success in these terms, the science park must succeed in a different dimension: economics. If companies, landowners, and the state do not reap economic benefits, there will be no science park. Paradoxically, however, too much attention to economic viability can conflict with authentic success. Under pressure to expand, a research park could easily evolve into a low-tech manufacturing or retail center, outwardly successful in commercial terms but making no contribution to the lab's (or the nation's) technology objectives.
Technology transfer. If the park grows as planned and becomes an economic success, the key question remains: Will it advance Sandia's missions? The first criterion for evaluating its success is technology transfer.
Starting with the Stevenson-Wydler Technology Innovation Act of 1980, legislation over the years has told the national laboratories quite explicitly that they must transfer commercially useful technology to the private sector and has given them a framework within which to accomplish this goal. Federal laboratories have worked hard to do so. From 1992 through 1995, more than 3,500 CRADAs were signed. Sandia has about 300 CRADAs, worth about $700 million in jointly funded research.
Sandia's initiative is based on the common-sense premise that technology transfer is easier the closer the user is to the laboratory. Although this is unquestionably true in the abstract, how important is it, particularly in light of the explosion of Internet communications? The evidence is auspicious but not completely decisive.
Although some forms of technology transfer, such as using patents or pulling data off the Web, have no transport cost, other forms do. Indeed, a great deal of technology transfer seems to require that people interact face to face. This is the point of research centers that bring together scientists who would otherwise be scattered around the globe. Conferences, professional societies, and personnel exchanges between institutions all take for granted that many forms of information transmission require personal contact. For technology transfer in its broadest context, research indicates that the most important mechanism is the movement of people back and forth between universities and industry.
Studies of how distance affects technology transfer deal mainly with the movement of university-developed technology to industry. The few relevant studies I have discovered affirm that proximity enhances technology transfer, and Sandia's own experience is consistent with this finding. Of Sandia's 300 CRADA partners, about half are located in New Mexico or the adjoining states of Texas, Arizona, and Colorado.
Another key premise is that there is a good market for Sandia technology. Among the national labs, Sandia has the research menu that is perhaps best suited to the needs of industry. Its strengths in materials research, fast computing, and semiconductor design are particularly well matched with current private sector activities. Moreover, Sandia's workforce is highly interdisciplinary, which makes its technological assets more widely sought after than those of a more specialized institution.
The science park model of technology transfer presents some new problems that Sandia will need to solve. For example, it will differ substantially from technology transfer under a CRADA in that S&T park firms will expect a steady flow of innovation, not just a one-shot technology transfer on a circumscribed subject. This will require staying ahead of the product cycle in fast-moving industries such as semiconductors. Moreover, to foster the rich and deep interaction required by a long-term collaboration, Sandia will need to ensure easy and frequent communication, unfettered by unnecessary restrictions.
Core missions. Can an S&T park filled with firms pursuing their own commercial interests contribute in any way to Sandia's core mission of national security? The answer lies with the productive synergy that so often emerges in well-conceived research and technology partnerships.
Ronald Kostoff of the Office of Naval Research writes about "heightened dual awareness," which is the understanding that often develops between researchers in fundamental science and the applications engineers with whom they interact. When these researchers operate in an "applications-aware" environment, their ideas tend to be naturally associated with applications germane to their research and hence to the mission of the federal laboratory. This effect has been documented in the well-known Project Hindsight analysis, produced by Battelle in 1973, and in a number of comparable studies since then. Kostoff argues that this heightened awareness of potential applications is necessary for productive fundamental research and that there are many ways to implement it. A functional science park could well be one of those implementations.
Technology transfer, it is now understood, is not simply a transfer of basic research to firms that turn it into technology. Rather, the process resembles a merger of complementary resources that together yield more than the sum of their separate products. One survey of company interactions with federal labs showed that although most of the projects involved lab-performed basic research, 39 percent involved company-performed basic research that could produce results of use to the labs. This same study found considerable evidence of two-way transfer of personnel, which we can assume produced a two-way flow of knowledge.
There are indirect benefits as well. The S&T park may make Sandia itself a more desirable workplace for scientists and engineers. For some, it will broaden the scope of their work. For those with an entrepreneurial bent, who see themselves eventually starting their own spinoff enterprise, the S&T park could be a convenient incubator. The park might help recruiting, because job opportunities for spouses at the S&T park could be an attraction for prospective employees.
But potential problems must not be ignored. As experience with the CRADA mechanism has shown, the technology transfer process is arduous. Typical CRADA problems would apply, such as the question of eligibility for foreign companies and complaints by nonpartners about alleged favoritism to firms that are partners. As difficult as a one-time agreement may be, the long-term and wide-ranging nature of an S&T park relationship is likely to be even more difficult to maintain.
One appealing characteristic of the Sandia S&T park is that the lab is not risking a large initial investment. As lab officials are quick to point out, their developmental costs are quite modest, consisting almost entirely of the staff time needed to manage the enterprise. Sandia subsidizes no one. The land will be sold or leased at market prices from DOE and local owners. Firms pay for construction. They may qualify for industrial revenue bonds or other concessions common to local economic development programs, but Sandia does not bear these costs. Sandia's subsequent costs involved in technology transfer, however, would have to be tabulated just as in traditional CRADAs and other such vehicles.
This means that the focus of evaluation can be on technology transfer and the contribution to Sandia's own missions. Whatever the approach taken to measuring these results, the evaluation system should be set up immediately, so that it can collect the needed data as they are generated, not years after the fact when such information may be unobtainable. A real-time evaluation system could also serve as a management tool to help focus the park's objectives. Finally, the evaluation results could at least shed light on the question of whether other national labs might use-fully emulate Sandia's initiative.
In the bigger picture, the Sandia S&T park should be viewed in the context of a rapidly evolving U.S. economy and a national laboratory system fighting for continued relevance and validation as part of the nation's science enterprise. Mere survival of the labs is not a national goal; contributing to the nation's security and prosperity is. If Sandia succeeds, it will have forged a new tool by which the national labs can help achieve these goals.
Kenneth M. Brown is a senior advisor for policy in the Directorate for Social, Behavioral, and Economic Sciences at the National Science Foundation.