-----Question-----
From: Ed Lazowska
Sent: Thursday, November 06, 2003 10:43 PM
To: phycos at u dot washington dot edu
Cc: Ed Lazowska
Subject: A question ...

Hi, Chuck.  Got a question for you ...

One of our first-year graduate students cornered me this morning with an
IP question.  Over the past year, he came up with an idea that might at
some point be commercializable.  He did some preliminary work on it, on
his own, verifying that there was some promise.  Then he dropped it to
come to graduate school.

The general area of the idea might also be a worthwhile thesis research
area -- not the specific application he has in mind, but the general
technique.

Here's what he's wondering:  Suppose a couple of years from now he wants
to try to commercialize this idea.  And suppose he does some research on
a closely related topic.  How would one determine what he owns and what
UW owns (and he must license)?  Is there any sense in him writing down
-- either for you or for the file -- the ideas that he is bringing into
UW with him?

He spent much of the past year consulting for a company, and when he
went to work for them, they had him list stuff that was in his head, and
he listed this -- common in employment agreements.  It's this that
brought the question to his mind -- should he list the stuff he's
bringing in his head to UW, so that there is on dispute regarding
ownership of the idea later?


-----Answer-----

From: Charles R. Williams, J.D., Ph.D.
Sent: Sunday, November 09, 2003 9:39 PM
To: Ed Lazowska
Subject: RE: A question ...

Hi Ed,

Sorry for the slow reply..got sidetracked with parenting.

Short Answer/Recommendation:

A student could place a sealed notarized statement in his/her student file
at CSE that will be opened and reviewed in the event that an innovation is
made on the idea that they think they brought with them.  At that time, we
will make a determination if the information is sufficient to rise to the
level of conception, and if necessary will consult with an outside patent
attorney to make sure we got it right.  As part of the student's file, this
gives the student the best protection with respect to disclosure at a public
institution because of FERPA and student privacy.

Additional thinking:

It is very common for companies to ask prospective employees to list prior
obligations and private IP in an employment agreement.  Smart practice.  We
face some different issues here that (of course) complicate the landscape.
But, there are definitely instances where a clear record of past work would
be very useful! 

Companies simply bar an employee from bringing in personal IP (or IP from a
prior employer).  The company would require rights in the disclosed work to
invest company resources in it (or bar the employee from working on it).
Universities are in general more flexible and don't force assignment of
rights (or exclusion of general ideas) on entry.  We typically make a
determination at a later date as to where the chips fall.  This is sometimes
painful and having accurate information regarding the history of development
would be very helpful!  BUT.the cost of administering a centrally managed IP
intake review system and the resulting review of relationships under the
watchful WA state ethics law (no use of state resources for a private gain)
would be problematic at best.

Further complication--Work under Federal Funding also brings in the
Bayh-Dole regulations.  Bayh-Dole brings university ownership into the
picture if the invention was conceived OR "reduced to practice" under the
grant.  Plenty of cases where something was conceived prior to work under
the grant but reduced to practice at the institution.  In these cases, the
university owns the invention.  Disclosure upfront won't change the
university ownership result where Federal dollars are involved in reducing
to practice.  Want to make sure incoming folks are not surprised by this
result.

It is definitely a good idea for an incoming graduate student (particularly
a student that has been working in industry) to discuss his/her past work
with their future supervisor or grad advisor.  As an example, we are helping
Brian Curless manage Dan Goldman's relationship with ILM. If there appears
to be some concern with respect to personal work or possible issues with a
former (or current) employer, it would be in everyone's interest for the
student to document the prior work/relationship.  If a student brings up a
concern or the faculty member spots a potential problem, the student can put
the notarized record in their file and we all then have something to go back
to if commercialization (or even just distribution) occurs.

The approach we have used to try and help manage IP rights more effectively
on a case by case basis is to have research projects setup their own
Participation Agreement that includes a section addressing possible
conflicts due to prior work/relationships and under what terms prior work
will be shared.  Individual PIs decide whether their research work on a
particular project requires the more formal rights management that a
Participation Agreement anticipates.  Usually the triggers to put a PA in
place are things like an industry sponsored relationship, proprietary data,
contributors with prior work that may be included in the research, or
collaborations that anticipate specific distribution paths for research
artifacts.  In the end, we often need to aggregate IP to put something
useful into play and distribute it.  Imagine a case where a grad student has
an idea they bring into UW and it is indeed patentable.  Lets say he/she
collaborates with a number of other students on elements of the method and
the other students are hard core GNU folks while the student is more
entrepreneurial.  The student files a patent based on his/her idea prior to
UW entry and the other students start to distribute code implementing the
patented method under a GNU license.  Under copyright law, authors can
non-exclusively distribute (need to share equally in financial gain). The
students have no rights under the patent though BUT GNU grants a license to
any patent that covers code. Ugly result for everyone.  

Interesting stuff.

Cheers,

Chuck

Charles R. Williams, J.D., Ph.D.
Director
Office of Software & Copyright Ventures
University of Washington
Box 354990
Seattle, WA 98195
(206) 616-3451
Fax (206) 616-3322
http://depts.washington.edu/ventures