Matthew O'Keefe

How to Co-Innovate with Universities

Blog Post created by Matthew O'Keefe on Jan 30, 2015

HDS has been leading an effort within Hitachi Ltd. group companies to

pursue open innovation between Hitachi teams and outside groups

including corporations, Universities and independent research labs.

These kind of collaborative research efforts can help Hitachi and its partners

improve their research outcomes and potentially get more of them to

market more quickly.


As someone who spent 10 years as a college schoolteacher and researcher,

I worked on several successful collaborative research projects between

my teams at the University of Minnesota (where I am still an Adjunct

Professor) and industry and other labs. In this blog post, I wanted to outline

some of the methods and principles we used.


[1] Universities consist of faculty who teach and students who are often

taking classes.

Although this sounds obvious, its important to remember that Universities

generally have very few full time research staff that can just focus on your

joint project. Faculty have a lot of other responsibilities including teaching,

administrative support, advising, and fundraising. Students have classes,

exams, and other responsibilities as well. Therefore, as you think about what

roles and responsibilities, remember that faculty and students generally

cannot focus 100% on your project, except potentially during the summer months.


[2] The University reward system is focused on published papers in the

highest quality journals, and the resulting prestige and general recognition

for this published research work.

Contrary to what you may have heard, University promotions are not generally

driven by how much money a particular faculty member brings in. Money is

viewed as a secondary indicator of research success. The primary indicator

is published research in outstanding journals in the field. In addition, graduate

students (both Master's and Ph.D. students) are expected to publish there

work, which requires that the work be innovative and relevant to the existing

body of work and peer-reviewed literature. This also makes signing NDA's

more difficult, though not impossible. University researchers are measured and

promoted based on what is published in the open.


[3] Bottom's-up and top-down: two different interaction methods with


The classic University can be thought of almost as a medieval monastery: each

monk (professor) has his own cell where he prays, writes, and performs his

assigned duties. Generally he is performing these duties wholly independently

of other monks, and he might have junior monks (students) assigned to work under him.

Occasionally the monks get together and sing, eat, and pray as a group, but

mostly they work on their own. Similarly, University faculty are independent

agents who typically define, resource, and execute their own projects. However,

this model started to change in the 1990s, where it was clear that the scale

needed for some projects (especially software projects) and more funding,

especially from agencies, was focused on group collaborations and getting

faculty to work together in Centers of Excellence via other kinds shared research

funding. This has now become much more accepted, and even

expected, and allows outside collaborators like Hitachi to leverage this

organizational infrastructure at Universities to work on larger projects than what

would be typically possible with a single faculty member.


[4] Keep the innovation Open.

Although Universities nowadays have mechanisms for keeping research secret

and for filing patents and establishing ownership of valuable IP (and they should!),

it can be difficult to work out a shared IP agreement in advance for joint work.

Universities get most of their funding from the government and, since 1980,

(when the Bayh-Dole Act was passed) effectively own the IP generated

from their government funding. It's quite likely that your joint collaboration

with the University is going to be part of an ongoing government-funded

research project for which the University already may own patents, copyrights,

and (less likely) trade secrets, so any effort to jointly or even exclusively own

funded work at the University may well clash with existing IP agreements in place.

So how do you avoid IP entanglements while still leveraging Universities to

achieve open innovation? Here are some suggestions:


(a) Break off a chunk of work for the University so that any IP they generate

works for you even if its open (e.g., open source software) and owned by the

University. For example, say Hitachi CRL (Central Research Lab) builds

an innovative new computing chip that has no programming language much less

software. Hitachi needs example software that shows off the chip's capabilities.

Let the University write and own the software programs and publish them

in open source, which helps encourage others to write software and helps

accelerate innovation towards a large base of software for the new chip.

Given that the University software will not likely be fully production-ready,

(but good enough for a research paper) nor can Universities generally

support and maintain software long-term (given the huge expense and distraction

from the original University mission of teaching and research), this is a win-win

for both parties.


(b) Allow and encourage whatever software is generated in the project

to be open source, even if the University owns the copyright. Open source

started in Universities (Stallman's GNU project and Berkeley UNIX are

where open source began) and its

a great fit because it allows open peer reviews, prevents software from

becoming stranded after students graduate and funding dries up (which it

inevitably does), and allows small University teams to leverage the open

source community to increase project scale and capabilties. Corporate

partners can even run with the University open source and find ways to

extend and commercialize it if they so desire.


(c) Define a project that allows the University to interact with your technology

at its boundaries where the proprietary IP is not visible. For example,

a project that involves testing your new computing device or hardware under

a specific workload that a University lab can uniquely generate would be

a good example of this. The lab doesn't need to know exactly how the device

works, just what the interfaces are and expected use cases. Here, the

corporate sponsor gets great feedback on its product without having to expose

any proprietary IP requiring NDA's or other complicated joint IP ownership



Universities are a fantastic resource for open innovation. The one thing

I learned about students is that youth is often an unparalleled time in your life

for innovative work because you are less weighed down with preconceived

notions. This diversity of viewpoints and swirling ideas

is a wonderful place to insert innovative corporate technologies that need

to be taken to the next level. With some understanding of both the strengths

and weaknesses of the University environment, both Universities and their

corporate partners can accelerate their innovation jointly.