Research- [ of | on | with ]

There’s one more important area of research invention use, and that is as a research tool. In the practice, we can differentiate three kinds of tool activity, in addition to throwing bones into the air and heading for monoliths orbiting Jupiter. These tool uses can be characterized as evaluation of, research on, and research with.

Evaluation-of is at the core of scholarship. It is how published claims get verified. If you claim a research result and that result depends on a new tool, such as a sensor, a data set, or a software code, then other scholars need to have access to evaluate the claim. This is where we get into the politics of science. If scholars only allow their friends access to their tools for evaluation, then the independence of peer review is seriously compromised. Patent folk worry about evaluation-of because it could come up with data that varies from the claims made in the original research, and that in turn might hurt the commercial marketing efforts or even patent prosecution.

Research-on has to do with exploring the operation of the tool beyond published claims. How can the tool be modified, extended, improved? Patent folks worry about this kind of thing, because research-on can result in improvement patents, can beat one’s own researchers to other aspects of a cluster of inventions. One may have reported a general effect but not have figured out the best composition of matter by which the effect is enhanced. Research-on worries point up the uniform lack of resources in the university community to deal with improvements scattered over a number of institutions. Each university asserts its provincial position and sits on its rights until forced to work out some sort of arrangement with another institution. In aggregate, the universities present a barrier to entry to anyone trying to put together a coherent package of rights. Worse if the package of rights is intended to be non-exclusive.

Research-with means the tool is useful for others conducting research. Research can be conducted for any number of purposes—to extend knowledge, yes, but also to develop new therapies and to stake out new areas for study. Patent folks go white knuckled at the thought of letting go of research-with rights. After all, some of the big hit winners in university licensing have been research with inventions–Cohen-Boyer (gene splicing), Hall (expression of polypeptides in yeast), Axel (cotransformation). And some of the most bothersome research issues foisted on universities by industry have involved research with tools–cre-lox and PCR, among others. Thus, in university IP admin thinking, research tools smack of product sales possibilities. The distinction between academic research and industry research doesn’t much matter. The academic research could be funded by industry. Or it could compete with one’s own institutional position for grants.

If we look at the CANVIS development for a minute, we can ask where these research practices end up. CANVIS stands for the five primary areas of university support for research results—commercialization, arbitrage, new venture, internal implementation, and scholarship. Commercialization means developing and selling product. Arbitrage means selling a right to others for less than one paid for it. New ventures are start ups, whether for-profit or non-profit. Internal use focuses on self-implementation, without the need for a product and without a specific interest in creating a product for others. Scholarship has to do with all aspects of scholarly communications, exchange, and review. University research policy has to deal with all five, not just one.

Evaluation-of certainly supports scholarship. We note that scholarship does not have to be restricted to work done at universities, or even at non-profit research organizations. Companies may have research labs that are every bit as scholarly as their non-profit cohorts. But evaluation is also an early step in commercialization and internal use discussions. While scholarship might benefit from open evaluation-of, efforts to market patent rights often require the use of a confidentiality agreement or even an option agreement before allowing access to invention materials and information. As one gets worrider and worrider about it, one worries whether a university could obtain evaluation-of information while working on a research project funded by an industry sponsor with rights to data and inventions. Is this paranoia or prudence or simply a failure to properly construct a research rights administration program?

A lot of the university research use world caved in with the Madey v. Duke decision. With that decision went most of the folk concept that there was a research exception within patent practice. The research exception operated a bit like fair use—that an invention could be studied and built upon as a way of promoting progress. It was a sort of “fair competition”. Perhaps it still is. But with Madey, companies—and universities—had to worry about infringement even in research settings. In practice, there are not a lot of incentives to go about pursuing research infringement, but in pharma situations, where billions of dollars may be riding on patent rights in a class of compounds, research-on restrictions still seem to be very much in play.

What ought to be happening here? First, universities should be re-implementing privately a broad research exemption. It is in their best interest. It is the role they should be playing. It is good public policy. This exemption should reach uniformly to all evaluation-of, regardless of the purpose or tax standing of the evaluator. Declare public purpose and make this a general license for all technology under administration. Go further and offer to help in that evaluation of by providing not only confirmation of rights but also data sets and expediting access to the lab reporting the invention. This access is especially important in the six months following a report of invention, where the experimental set up is likely still in place and others can not only learn about the invention but observe the experimental conditions under which work has been performed.

Second, universities need to develop administrative tools to support research-on. Under Bayh-Dole, there is a mandate to administer patent rights so that further research and discovery are not impaired. Research-on indeed represents a direct pathway to improvements and springboard inventions (which may not improve the original invention but are based on it). Restricting research-on work by requiring paperwork or making demands for access to resulting IP (“reach-through” requirements) restrict both rapid advancement and competition in research. Typically, universities in granting exclusive licenses reserve rights for research and educational uses. These reservations, however, are typically only formalized at the time of the exclusive license and are never announced publicly. This practice needs to change. Universities should announce these reservations of rights publicly when an invention is accepted for management. The reservation of rights—essentially a general public license for research-on—should be included in all non-confidential summaries and in publication acknowledgments, so everyone is on notice.

The challenge with research-on is gathering back together fragmented rights. One way to do this is with a research consortium, patent pool, or research commons. Another way is to create an open exchange, in which rights are documented and made available to be re-mixed without a lot of bi-lateral negotiation delay. Critical to any of this are solutions to two key issues; royalty-stacking and point of control. Bayh-Dole made uniform federal agency policies on invention procurement in federally supported research. Now it is up to universities to create uniform protocols for providing access to industry and the broader community.

We have to recognize that most university research originated technologies involve multiple inventions, multiple IP, multiple organizations. This is the default condition. Yet university IP policy and practice starts with solo case, solo IP, and solo organization, and then tries (or doesn’t even try) to extend to the case of multiple. The default is itself a flaw in seeking collaboration. Provincialistic over-simplicity is the default. Multiple should be the default. Solo is a strange case with a lot of warning signs. Absurdly simple = danger.

Research-with represents the most challenges. Tools can be products. There can be commercial markets for such tools, even if a lot of the buyers are no n-profits. Here, the move has to be made into support for the internal use audience. Technology transfer is teaching with enablement across a new technology in context of the infrastructure and tacit knowledge needed to implement. Not everything is taught via productization.

Universities need to separate management of internal-rights licensing from commercial product licensing. A sound measure of the value of a product version of a research invention is that it adds value over what organizations can do for themselves. If an organization can self-implement by reading publications, then using a patent right to keep them from doing this is counterproductive. Doing so does not promote use, but rather suppresses use in favor of a monopoly position. This may be the money position but it is lousy headed public policy. The basic way to do this is to separate make/use rights from make/use/sell rights. If granting broad make/use rights would prevent any company from wanting a sell right, then right there one can question the value of the patent as an asset for research administration.