Research Teams at Universities: The Center for Interfacial Engineering
D. Fennell Evans and Matthew V. Tirrell
University of Minnesota
The Center for Interfacial Engineering (CIE) was funded by the National Science Foundation (NSF) as an Engineering Research Center (ERC) in October 1988. There are currently 26 ERCs in the United States, each of which is funded at an annual level of between $1.5 million and $3.0 million for up to 11 years. Most of these centers focus on a specific technology (for example, electro-optics, deep-sea platforms, plasma processing, and so forth). The CIE is somewhat different; our goal was to develop a new cross-disciplinary engineering field: interfacial engineering.
Competition to obtain an NSF ERC grant is extremely fierce. In the year we applied, there were 68 proposals, only 4 of which were successful. Our strategy was to suggest that establishing a new engineering field focused on interfaces addressed a pressing need for U.S. competitiveness. Our commitment was that industry would be a full-fledged partner in this endeavor. Our proposal drew on existing strengths in the University of Minnesota's Engineering and Science departments, included letters containing substantial financial commitments from 20 U.S. companies, and contained a major financial matching commitment from the University of Minnesota.
After 18 months and several site visits, we were finally funded, but only at two-thirds of the proposed level. Our first difficult task was to narrow the scope of the center's activities to match our resources. We decreased by half the number of faculty involved in the center and redlined a number of proposed programs. There is no activity in a university that is so virtuous that it should be undertaken in the absence of the required resources!
The original vision for CIE, which remains unchanged, is to
- Lead the development of a fundamental understanding of interfacial processes,
- Establish synergistic transfer of technology between CIE and industrial partners, and
- Produce well-educated graduates who can apply their knowledge of interfacial operations to new processes in a productive manner.
The expectation is that the research program will be first rate, addressing issues that will impact national competitiveness and will be carried out in direct collaboration with industrial researchers. The
major driving force behind the technology transfer program is that it is proactive, managed by an individual with industrial experience, located in the center, and not managed as a adjunct activity by lawyers hidden away in one of the university's many bureaucracies. The expectation of the education program is that it goes beyond the traditional university programs by introducing students to the systems approach to projects. More recently, the production of exemplary instructional materials has been incorporated into the NSF's ERC expectations.
All ERCs are exhaustively and seemingly continuously evaluated. The NSF performs an annual site visit with five to seven external reviewers, who typically produce a 7- to 14-page review. Critical reviews for continuance occur during years three and six. Several ERCs have been terminated as a result of an unsatisfactory third-year review.
In addition, the agreement with the NSF calls for an annual report containing substantial amounts of detailed data that consumes the equivalent of one full-time staff person per year to compile. Because an ERC's operation involves handling millions of dollars from the federal government, the University of Minnesota, and member companies, the activity attracts auditors like vultures to prey. Dealing with the expectations of companies and simultaneously with the rules and regulations of two major bureaucratic organizations requires a sophisticated and highly skilled staff. The costs of running an ERC properly are not trivial.
Organization and Operation
As shown in the organizational chart in Figure 4.1, CIE has four research programs, each under the leadership of a professor. Associated with each of these programs is a technical advisory committee that meets twice a year with faculty to review progress and suggest new research challenges and directions. All research projects at the CIE are nonproprietary in nature.
A major focus in the center has been the development of a Characterization Facility that currently contains over $7 million of high-level research instrumentation. This facility has become a major research resource for the university and was used last year by 400 university and faculty personnel and by 120 companies. The annual operating budget of the facility is $700,000 and is obtained from instrument-user fees. These funds cover the salaries of the facility personnel and service contracts on the equipment.
As part of establishing a new discipline, the center sponsored the publication of six textbooks in a series entitled ''Advances in Interfacial Engineering."1 More than 10,000 copies of these books have been sold. In addition, the center has developed several series of computer-based instructional modules in interfacial engineering, fluid mechanics, and thermodynamics. The university has licensed these to a private company for distribution.
The magnitude of the operation is perhaps best depicted in terms of the statistics applicable to our recently completed tenth year of operation (see Box 4.1).
There are two main membership categories:
- Sponsors, who pay $80,000 per year and are involved in all of the research programs, and
- Affiliates, who pay 0.03 percent of annual sales with a cap of $30,000 and are involved in only one of the research programs.
Company involvement in the center is governed by a standard membership agreement between the university and the company. Obtaining an acceptable agreement is a not trivial task. We spent our first 18 months negotiating an agreement with member companies. Membership dues are paid into the University of Minnesota Foundation free of indirect costs.
According to the membership agreement, sponsors can license all intellectual property developed in the center. Affiliates are limited to developments in the program of which they are members.
A major activity in the CIE's technology transfer program is the Industrial Fellows Program. The origin of this program was the realization that many forms of technology transfer (i.e., brief one-day meetings, annual reports, reprints, etc.) are not very effective. The obsession on licensing of patents is often counterproductive. With few exceptions, most universities spend more on legal fees associated with patenting and licensing than they take in from the process. To allow effective technology transfer,
BOX 4.1 Statistics for Operation of the Center for Interfacial Engineering, Year 10
sponsors can send an employee as an Industrial Fellow to the center for three months or longer and receive a $50,000 rebate on their dues. These fellows work on a generic research project that has been negotiated between the center and the company. More than 140 Industrial Fellows have been involved over the past ten years. Most of the fellows continue to remain involved in center activities after their residency is concluded. In fact, there is even an Industrial Fellows alumni association. These Industrial Fellows have operated as effective change agents in our university environment. In addition, whenever they change employers they become very effective recruiters by having their new company join CIE.
Getting all of the legal agreements in place is an enabling step, not an end in itself. Our premise is that companies offer the university every bit as much as the university offers companies. Translating this belief into realization requires driving a cultural change that, if successful, leads to the establishment of a true partnership between university and company researchers. To be blunt about it, part of the challenge is to get the faculty to stop lecturing to the company researchers (i.e., to shut up and listen). The other part is to get the industrial people to talk. That seems to take about three years.
Developing strategies to sustain the dialogue between companies and the center constitutes a second major challenge. Over the years, we have developed a joint skill of translating industrial-based concerns
into high-quality Ph.D. theses. The technology transfer operation plays a vital role in this activity and becomes the formal exchange point sustaining continual, more informal exchanges between researchers. We keep a log of important industrial conversations and action items; our log now contains more than 14,000 entries.
From the company perspective, it is essential that all action items that come to the attention of the center are followed up and acted upon. From the perspective of the center, it is important that commitments to companies are implemented. Having a Technology Transfer Office with extensive industrial experience and whose native language is "industrialese" is essential. We professors speak "professorese," and a translator is a necessary ingredient to building an effective and sustainable partnership with companies.
One of the hallmarks of an ERC operation is the development of a strategic plan to guide the development of the center's program and provide benchmarks against which to measure progress. A number of the achievements are summarized as follows:
- Accelerated the development of interfacial engineering as a new interdisciplinary field;
- Graduated 522 students with center experience;
- Established the Characterization Facility via a $7 million equipment investment;
- Delivered short courses and workshops to companies;
- Hosted some 123 company researchers on campus as Industrial Fellows; and
- Developed textbooks, multimedia instructional materials, courses, and practical experiences for students.
Maintaining the Industrial-University Partnership after Engineering Research Center Funding
At the beginning of the center's operation, we set into motion a strategic plan for sustaining the center after the end of ERC funding which, at that point, was 11 years away. This initially involved careful management of cash flows and interest on the University of Minnesota Foundation funds to generate a substantial reserve fund. More recently it has involved careful planning with the center's Planning and Policy Board as it becomes the main stakeholder in the center's operation.
David Schetter, University of California, Irvine: Were patents exclusive to companies outside the consortia? Was there a university patent agreement?
Matthew Tirrell: Members of the consortia have a right of first refusal for a most-favored, nonexclusive license of patents on inventions made solely by University of Minnesota faculty, staff, and students. If no member company elects to obtain a license, then the patent reverts to the University of Minnesota' s Patent and Trademark Marketing operation to handle. This is a quid pro quo for the University of Minnesota's support for the center. All member participation agreements include the university's standard patent agreement.
Andrew Kaldor, Exxon: Was the measure of success for the Industrial Fellows the repetitive business?
Matthew Tirrell: Yes, the principal measure of success was the extent to which members continued to make cash contributions and remain involved, in spite of their own cash-flow problems, reorganizations, and difficulties in designating Industrial Fellows.
David Schetter: Was it really an award without university overhead? Was it really a gift?
Matthew Tirrell: Yes, in accordance with University of Minnesota's policies, these industry contributions were considered as unrestricted gifts.
Richard Alkire, University of Illinois at Urbana-Champaign: Is there just the one category of membership?
Matthew Tirrell: There were two major categories of membership: the sponsor, contributing $80,000 per year, and the affiliate, contributing $30,000 per year. Sponsors can participate in any and all research program areas, whereas affiliates can participate in only one.
James Desveaux, University of California, Los Angeles: You have outlined an impressive degree of innovation and adaptation. A question by another member of the audience has spawned curiosity for me about some organizational design issues, related to the different types of industrial partnerships permitted under the arrangements you described. What you have described sounds a lot like the kind of system that the Microelectronics and Computer Technology Corporation in Austin put in place a number of years ago. In fact, what you have discussed even bears some resemblance to partnership innovations recently made by the partners of the Semiconductor Research Corporation. Where did the ideas come from for the collaborative innovations that you outlined? Did you actively scan other models? To what extent are these models replicable?
Matthew Tirrell: The leadership for developing the proposal to NSF was shared by D. Fennell Evans (who became the center' s director following the award) and Robert Stokes (who retired from Honeywell after serving over 25 years as a research scientist and who also served two years as an NSF program officer). Evans and Stokes visited a dozen or more companies and solicited their advice and suggestions. These were incorporated into the ideas developed by key faculty members, who later became research program leaders. Although we were well aware of the NSF' s ERC guidelines, we "did it our way" for the most part.
Richard Alkire: What are some examples of how you changed what you are doing? You talked about it in positive terms, but I am curious about specifics.
Matthew Tirrell: This was an evolutionary process in which people and ways of doing things changed in response to opportunities that developed as we interacted with companies, especially through their Industrial Fellows. We soon learned that we could base high-quality Ph.D. theses on industrial problems based on a continuous dialogue between our faculty and students and their industrial counterparts.
Laren Tolbert, Georgia Institute of Technology: I've probably been an administrator too long, which involves indirect costs. You made the comment about faculty supporting all the administrators. But my administration would agree that these are real costs, which they are. William Wakeham talked about a plan in which the full costs were paid, although you never said exactly what your indirect cost rate is.
Are we selling our research too cheap when all we ask for is the dollar that goes to the student and not the dollar that pays the lights or pays the person who writes the check to the student? Therefore, are we subsidizing the industrial research through resources that we have?
Matthew Tirrell: Unfortunately, the typical university perception is that academia has a lot to offer companies, while companies have nothing to offer in return. We established the CIE as a partnership that brought tremendous advantages to the university. These advantages will pay dividends for a long time to come.
Bruce Harrer, Pacific Northwest National Laboratory: I had a follow-on question to that with regard to the indirect costs. There are a lot of centers listed in Figure 4.1. Did they all get a waiver on their indirect costs? Also, for the source of funds in Figure 4.1, is it only the industrial contribution that received a waiver on the indirects, or did all these principal investigator grants get a waiver on the indirect costs as well? Eventually, someone has to pay for the overhead for these centers, and if you grant waivers, you will have to pay for it somewhere else.
Matthew Tirrell: Indirect costs were paid on the NSF ERC grant and on all grants to principal investigators. The ERC and other grants dictate a large number of nonresearch activities (e.g., time-and effort cards) that companies have no interest in paying for.
Christopher Hill, George Mason University: As you've gone from the original center to the new center, have you noticed a difference in how industry is interested or willing to relate to you? It' s getting tougher to get this kind of open-ended, nonspecific support. Is that your experience?
Matthew Tirrell: Upon the receipt of the award for the new Material Research Science and Engineering Center (MRSEC), we structured, in concert with our member companies, the "CIE/MRSEC Industrial Partnership" which allows companies to participate in the activities of either or both centers by executing a standard participation agreement. In their other relationships, the two centers operate independently.
David Schetter: Have you approached your current members for sustaining or increasing any contribution to give you a core; and if you have, what was their response?
Matthew Tirrell: Over the past ten years, there have been two increases in annual fees, agreed to by member companies. As we continue with the new industrial partnership, the same basic fee structure will apply.
Thomas Manuel, Council for Chemical Research, Inc.: To link Christopher Hill's point with Laren Tolbert' s for this discussion, we are dealing with a classic perception in academia that things are priced based on the cost for them. But this is not the case. Things are priced based on their value in the marketplace. And companies, due to the dynamics of the globe today, are forced to be very hard-nosed about their own profitability, getting the results for their shareholders at the lowest possible cost.
So the issue is not that you are selling your research too cheaply. The issue is whether a company will pay whatever you charge compared with what, for example, Imperial College, Taiwan, or what a cooperative research and development agreement involving Los Alamos will charge. This is a common disconnect between the academic and the industrial communities on this issue, and we need to be aware of it.
Matthew Tirrell: At CIE, we have been careful in setting the prices we charge for what we offer and have always done this in consultation with our industrial advisory board. Also, our member companies report great benefit that has accrued by their taking the results of our nonproprietary research and applying them in the context of their own products and processes. We have solicited "nuggets" from companies regarding the benefits from their participation in the center. In one case, the company reported savings of over $25 million thanks to the application of a technique learned by one of its Industrial Fellows during a campus residency. In another case, a new product resulted from the application of fundamentals garnered at the center. In all, we have recorded over 365 of these "nuggets."
Thomas Manuel: The problem is that the economics that I described are localized free market parochial microeconomics. In a macroeconomics sense, and in the enterprise, the nation has a different set of values. The timeframes are so long that individual companies find it hard to deal with them.
Matthew Tirrell: Indeed, this is an issue that we often confront. Some companies have simply chosen not to participate based on timeframe concerns. Some companies have chosen to withdraw based on the shortening of their time horizons. Simply stated, companies and their timeframes change quite often, with some companies coming and going.