Driving IT Innovation and Network Innovation
From early mobility projects to high-bandwidth research networks, IT--and Tracy Futhey--move the campus forward at Duke.
As CIO and vice president for Information Technology at Duke University
(NC), Tracy Futhey oversees technology projects that have far-reaching impact
on instruction, research, and the way people live and work on campus. While
recently she is most often cited for her work with Duke’s iPod project
(www.duke.edu/ipod),
dubbed the iPod First Year Initiative, Futhey’s professional interests
have spanned a range of innovative technologies, especially networked and mobile
technologies. At Carnegie Mellon University (PA) in the ’90s,
her projects included work on Wireless Andrew, an early campuswide wireless
deployment, and on Handheld Andrew, which foreshadowed some of today’s
emerging network and location-based services. Today, in addition to her role
at Duke, she is board chair for National LambdaRail (www.nlr.net),
the national optical research network. Her overall goal to “create dynamic
research and learning environments” is met by innovation and enthusiasm
both at Duke and at peer institutions across the country.
How have networked environments at colleges and universities changed
over the years, and what are some of the biggest opportunities that have opened
up? The biggest opportunity relates to broad access to these networks
and the experimentation that has become possible as the networks have moved
from being specialized—only available to a few— to true commodity
networks. I was involved in the Wireless Andrew project at Carnegie Mellon University
[nearly 10 years ago]. At that time, the idea of 802.11 wireless and what one
would now consider to be broadband wireless access, was something that people
in the mainstream simply didn’t recognize as having the potential that
it eventually turned out to have. Then, it was really a tough sell to get people
outside of the research community, and outside of some of the student community
at CMU, to realize that being able to walk around with your laptop and be connected
by the network was extremely powerful. Many had been using computers in a network
desktop environment and thought that was perfectly adequate.
So the networked environment itself encourages experiment?
The networked environment is an enabling capability that pushes people to try
new things. There are new things that one might be compelled to do because of
that mobility. I think of location-based services and context-aware applications—an
entire applications area only emerging as a result of the mobility that’s
being created by networked environments and the pervasiveness of networking.
The more you’ve got broad access and experimentation through this commodity
network—be it 802.11 wireless, or cellular and wide-area networking—and
the fact that so many people are being connected so easily and inexpensively,
without much effort, leads to a lot of individual situations where we can try
new applications. These new uses are being tried by individuals who are not
necessarily developers, but may be people who have a different perspective and
a great idea worth trying. So, many of the changes I see relate not only to
expanding access, but also to expanding the base of people who are trying to
do different things because they are now enabled through the networked environment.
Who drives the experimentation? Students as customers, innovators in
the IT department, or…? I would characterize that as having three
major drivers, if one is to be successful in running a strong and forward-looking
IT organization. One of those drivers would come from the students. And, increasingly,
the students are coming in with more and more experience in using technology,
and huge expectations about what technology can do and about access in virtually
every aspect of their lives. The uses they’re looking for when they start
out are primarily convenience uses: They are used to buying books from Amazon,
and paying bills online. So we have to think, how do we make the campus operate
efficiently and effectively in an online environment? And students’ willingness
to adopt new technologies is a critical aspect of that first driver.
You mentioned the IT organization and its innovation in pushing ahead. That
is the second leg of an important support structure. But a third one you hadn’t
mentioned, that is absolutely critical, is the push and drive that needs to
be coming from the faculty. And that relates to the use and support of technology
both in the classroom and in the research environment. It’s the one that
in many ways has driven us at Duke: Last year, with our iPod First Year Experience,
and in the coming year, with the Duke Digital Initiative; our program to continue
the iPod use and expand into other technologies.
Is there a fourth driver that you characterize as strategic?
It’s associated with the second driver, although not exclusively. Duke
has, as part of its academic strategic plan, a goal and a commitment to utilize
technology throughout all aspects of campus life and the campus environment—be
that in the classroom, in the research setting, or in the co-curricular aspects
of student life. I view that as an institutional driver, one that I have to
be constantly mindful of; one that I have to encourage and enable through the
IT environment, providing support for both the faculty and the students. But
whether you count three or four legs, they are all certainly necessary.
How is the iPod project expanding into the Duke Digital Initiative?
Over the past year, the iPod project has made clear to many of our faculty and
in many of our courses, the value of digital audio in a whole host of domains.
And those who have started to use the iPods have said, in a large number of
cases, that this has been an important enough experiment that they absolutely
need and want to continue it. And we’ve had interest from other courses
that haven’t yet tried it.
At the same time, part of what our evaluations identified last year is that,
as great as the iPod experiment was for digital audio access, we had people
who wanted to do things that aren’t confined to digital audio: digital
video, tablet computing, collaborative areas. So we are continuing and expanding
the project, adding the focus in those new areas. We are also shifting the project
from class-based to coursebased, so that any student iation?n any year of his
or her undergraduate program here at Duke, who is in a course using the iPods,
would receive an iPod for that relevant coursework.
Please tell us about national and regional networking initiatives,
specifically NLR and Internet2: What are the latest innovations, and what will
keep momentum moving toward building services for all institutions?
We’re making wonderful progress on the national, regional, and campus
level, to start to take advantage of the newly available optical networking
capabilities. We’re moving from the promise and the dream to the reality
of having very high-speed access from end to end—one campus to another,
one researcher to another— that was historically only practical to expect
within campus lab environments. The projects that are going on now at the national
level include NLR and Internet2. I2 has a project called HOPI, hybrid optical
packet infrastructure, that is using NLR infrastructure. That and other projects
at the national level are going to be introducing a whole new set of capabilities
for faculty researchers. The regional build-out of optical networking has really
been impressive over the last couple of years.
For me, one of the key elements in participating and moving those initiatives
forward is providing faculty on our campuses with the best access to research
capabilities and the fewest barriers to collaboration as possible. It is often
the case that collaboration within a particular discipline occurs across campus
boundaries rather than within a single department. So, through the high-speed
optical capabilities that we’re trying to introduce, we can make sure
that a faculty member at Duke, for example, can communicate and network with
research faculty at the University of California- San Diego
site, the other supercomputing centers, or you name it. I’m able to make
sure that faculty have such capabilities on this campus, and to collaborate
with their colleagues at other campuses, without regard for the fact that Durham,
NC, where Duke is located, has historically not been viewed as the networking
center of the world.
What can be done with grid computing, to improve research computing
environments for our campuses? Grid and cluster computing are having
an impact, not only within the research lab; they are also providing important
opportunities for faculty to collaborate across disciplines. If we do it right,
we’ll create opportunities for computing organizations to have a rejuvenated
research computing support environment for faculty.
Cluster computing, for example, takes significant effort and system management
to support; in some cases, faculty don’t want to do that themselves. So,
as a central computing organization, figuring out how we can support our faculty,
and creating flexible models to do that, creates an important opportunity for
faculty and IT department interaction.
You’ve been involved in significant network innovations. What
have you learned about the nature of innovation? The most critical
thing is tied to relationships with faculty. In general in universities, in
running our computing organizations, we don’t take as great advantage
as we might of our faculty’s experience and experimentation. In terms
of a strategy, I’m very big on the notion of working with faculty; meeting
with faculty in your computer science and engineering departments, in your school
of medicine, in art, in any domain, and understanding what capabilities they
need or may be creating. That’s the key to innovation: viewing technology
not simply as infrastructure, but as the gateway that can help provide faculty
and students with advanced capabilities. It’s not one-size-fits-all. There’s
got to be collaboration and connection with the academic units, and with the
faculty in particular, to understand capabilities and needs, and design around
them.