Number 31, December 2001

Guest Editorial

Industry-Academic Partnerships: The Oklahoma Experience

During the Reagan presidency, Eric Bloch, then director of the National Science Foundation, established the Science and Technology Centers (STC) Program in response to concern about an apparent weakening of the United States in science and technology development, as well as a desire to broaden university linkages with industry and develop more direct ties between university research and K-12 education. To be located exclusively at universities, the STCs were to attack especially challenging problems in basic research that required relatively large, interdisciplinary teams and stable funding for a significant period of time -- problems that eventually might yield technology for the benefit of society. Further, the STCs were to develop meaningful relationships with industry for effectuating knowledge and technology transfer; provide a framework for improving the representation of women and minorities in science; and develop programs for linking university research with K-12 education.

In 1987, Doug Lilly and I responded to the first STC solicitation to create the Center for Analysis and Prediction of Storms (CAPS) at the University of Oklahoma (OU). We proposed to study the practical predictability of high-impact, local weather, especially deep convective storms, using numerical models initialized principally with data from the NEXRAD Doppler radar network. At that time, no one knew if small-scale, highly energetic weather phenomena had any degree of practical predictability. However, we faced an even greater challenge -- one that formed the core scientific problem upon which CAPS was founded: the estimation, from time series of the radial wind and reflectivity fields provided by a single Doppler radar, of the dozen or so quantities needed to initialize a computer forecast model.

To address this and many other issues, CAPS built its foundation around the following principles, some of which immediately were evident while others emerged over time: 1) in addition to conducting research, develop a tool or technology (in the case of CAPS, an end-to-end forecast system) that will provide sustainable R&D opportunity beyond the formal life of the center, particularly in the context of operational forecasting and commercial applications; 2) as a national center, undertake initiatives that have broad community involvement and impact; 3) work with end users, especially in government and private industry, in evolving the science and technology; 4) establish close working relationships with other disciplines and focus not only on the meteorology of the storms-scale prediction problem, but also on the computational science aspects; 5) make all results and software publicly available for use without restriction; and 6) ensure that the benefits of the research have a clear and effective path to practical utilization.

When CAPS began operation in the late 1980s, Doug and I saw the National Weather Service as its primary, and perhaps only (in addition to the military) beneficiary. However, by the mid 1990s, that simplistic view had changed dramatically to include broad elements of the private sector. Today, the provision of meteorological services is undergoing a profound change, with the private sector emerging as a dominant infrastructure for the creation as well as delivery of customized weather and climate information. This transformation is being driven by four developments: 1) the availability to private companies of extremely powerful computers that, until recently, were affordable only to major government facilities; 2) the general availability of sophisticated numerical weather and climate prediction software, like that developed at CAPS; 3) a growing recognition by industry that access to customized weather and climate information, appropriately applied, can lead to significant economic advantage; and 4) tremendous advances in communications and related technology. A related factor continues to be the slow and cumbersome process by which technology moves from the research to the government operational sector, and the inability of the government to respond rapidly to changes in technology -- a response that, in the private sector, often can be nearly instantaneous.

All of these factors are leading private meteorological companies, and even large weather-sensitive industries such as energy and transportation, to operate their own numerical forecasting systems (which, importantly, continue to depend upon data from the National Weather Service). The specific configuration of this technology -- in contrast to the National Weather Service -- can be customized for particular decision making processes or tools. Thus, instead of simply repackaging NWS information, private companies now are generating independent forecast products on a schedule, and in a configuration, appropriate to their needs. Perhaps most importantly, however, is the fact that the user of the information generated can change the configuration of the forecast system to respond to the particular "problem of the day or hour." No such feedback loop exists between information providers and users in government-operated systems.

In line with Congressional expectations placed upon the NSF (and especially the STCs), CAPS began to emphasize technology transfer in 1996. At that time, American Airlines funded a 3-year, $1M R&D grant to further develop and apply experimentally the CAPS technology to its domestic route system. Not surprisingly, this level of funding brought an expectation by American of owning the intellectual property. Prior to 1980, such ownership would have been unattractive, for any technology developed with federal funds was at that time "community property." Arguing that what belonged to everyone belonged to no one because a company would not attempt to commercialize something to which it could not gain exclusive access, in 1980 Congress passed the Bayh-Dole Act (P.L. 96-517) to provide universities the option of taking exclusive title to intellectual property developed with federal dollars for the purposes of commercialization. In the case of American Airlines, OU -- like most of its counterparts -- chose to retain ownership of the intellectual property, but to grant American Airlines a license for its use.

Given that companies like American Airlines wish to use technology operationally, and given that universities are not private companies capable of providing guaranteed deliverables, services, or quality of service, CAPS led the establishment in Norman of a for-profit private company, Weather Decision Technologies (WDT), one goal of which is to commercialize intellectual property developed at OU. Like most other universities involved in spin-off companies, OU owns a small portion of WDT. Such linkages, though legal, raise legitimate issues regarding conflict of interest, which always will exist but which at OU are fully disclosed and carefully managed.

The successful partnership between CAPS and American Airlines, which formally ended in 1999 with the donation of $250,000 to establish at OU the American Airlines Professorship in Meteorology, set the stage for a $10.6M alliance between OU and Williams Energy Marketing and Trading headquartered in Tulsa, Oklahoma. As part of this alliance, which began in July 2001, CAPS is receiving $8.1M over 5 years for basic and applied research in numerical weather prediction, and for the purchase of a new supercomputer. The Williams project affords CAPS an opportunity to continue its work at a funding level equal to that provided by the NSF during its tenure as an STC, with additional grants and contracts nearly doubling that amount.

The intellectual property produced under the Williams funding also will be owned by OU but licensed in various ways to Williams. This protects Williams' investment without compromising academic freedom. No restrictions are placed on publications, including dissertations and theses, though of course CAPS will be careful to avoid disclosing within them information that is proprietary to Williams. Furthermore, all software developed can be used for non-commercial research and development, which of course yields benefit to the broader research community.

Developing meaningful relationships with private industry is a time-consuming and lengthy process, as are all relationships in life. Some worry that academic freedom is or soon will be compromised, and I too share this concern. However, based on my experiences to date, I believe that academic-industry partnerships and alliances indeed can work to the benefit of all, and moreover, I feel they are vital for economic development. The only way to know for certain is to try.

Kelvin K. Droegemeier
Director, Center for Analysis and Prediction of Storms
University of Oklahoma
kkd@ou.edu