For researchers in applied sciences, it is the norm to file intellectual property (IP). This practice can be construed to make a lot of sense:

• The objective of applied research is applications, solving real-world problems.
• Applications generally make it out into the real world because a company thinks they can make a profit solving a real-world problem.
• This company needs to make investments to develop an academic invention into a product and make a profit; they would like some leeway before the competition catch up.

This is what a patent can do: it buys time. The invention is made public, but barred from commercial exploitation to anyone except the licensee  of the patent, for about 15 years. With the proviso of a good patent, the applied scientist with a good idea can transfer an invention to a company, which will allow it to solve real-world problems. Applied scientists like that. S/he will receive license fees and royalties, which can be used for graduate students’ salaries and maybe even a bit of extra salary for the inventor. If s/he would have published the invention in a scientific journal, it would be in the public domain and everyone could use it, so nobody would use it.

So much for the theory. I am a newbie at the game of patents-for-academics, and that vantage point allows for healthy skepticism. I am discovering some aspects to patenting that make me reconsider the clearcut story outlined above. In the following I will allude to anecdotal and word-of-mouth evidence, the use of which I did not verify with the respective sources, so I will leave out names and other specifics. Email me for details. In no particular order:

Patents are a huge time sink and they are expensive: Seeing a patent all the way through costs roughly 40-50 k€. How do you justify this kind of expense? The only solution is to license it away early (i.e. way before the patent is granted). The time spent on a patent is estimated to be about three times as much as on a full paper. As Ad outlined in an earlier post, it takes a long time (up to 6 years) before a patent is granted. And the farther you get into the process, the more difficult/detailed/nasty the questions get. The inventor, you, is the one to answer them. Can you replicate the intricate details of experiments from 6 years ago? Everything needs to be documented, including seemingly tangent thoughts and arguments pro and con. This is a lot of work.

Patents do not count towards your publication statistics, even if your own institute recognizes them as valid output: Researchers are judged on their peer reviewed publications. Patents are often included in a publication list (all the way at the end), but are currently not included in the literature databases. This is where a grant jury will look for your track record.

A university is not a good place to manage patent revenues, even if your own institute recognizes them as valid output: Universities are always strapped for cash. If a research department has some of it (e.g. from patent revenues), they will try to take it away. Really. A highly successful research group at a Dutch university had an invention that was successfully licensed to a large company, and turned out to be a game changer in the industry. At the peak of activity, the group had 30 researchers working on the effort, most on patent licensing money. Eventually, the university budget for the group was cut because (a) they had enough funding anyway and (b) they did not publish sufficient peer-reviewed articles in the years prior — no, they patented their work!

Patenting is a long term process, academic research can not guarantee the continuity required: A university professor with a post-doc invents something, possibly useful. They talk to their technology transfer office (TTO), apply for a patent, which takes years. Post-doc moves on, maybe even professor moves on, at least thematically, and the TTO is still trying to license away the IP. It is a single patent, not part of a bigger portfolio, there may be workarounds, technology moves on. All the inside, nuts-and-bolts knowledge has disappeared. It is impossible for a company to build on IP in that condition. Cost for the university: 50 k$.

As an academic, you are in no position to evaluate the work of the patent attorney. And not all of them are equally good at everything they do. (See “good” patent emphasis above.) A patent needs to fulfil certain criteria, which are carefully hidden in a bunch of legalese by the time the application gets filed. Only after 6 years when it is rejected, will you find out that e.g. the claims are not supported by the text, an easy workaround was not blocked, etc. Because the actual text is written by a patent attorney, there is no way to fully control this.

Companies hoard IP, not for innovation but to keep it from their competitors: Knowledge gets shelved, enforcing product development is impossible. This is the most frustrating situation you can be in as an inventor. You’d wish you’d published it.

The above arguments spell out why patents in an academic setting may not work, or may not work for the inventor putting in effort where s/he could be writing papers. Of course, there are examples of academic inventions solving real-world problems. A few conditions, or at least helpful circumstances, for this to happen are:

• The company acquiring the license should be involved (under NDA) at the moment of filing.
• The invention should cater to a niche market. If there is too much money to be made, the patent will go under in litigations.
• Licensing the IP is better than selling because there is a (yearly) recurring cost. The IP department needs to justify the license fee to the board every year. If the company does not use the IP, they return it, and it can be licensed to someone else.

Related posts:

1. How difficult is it to get a patent?
2. Doing multidisciplinary research

E-mail    •

Share:

Tweet