Beyond Networking: Mobile Computing on Campus
The wireless networks that connect us are becoming
more pervasive, and digitized information can be accessed by a wide range of
mobile devices. Here, Syllabus talks with Charles R. Bartel, Carnegie Mellon
University’s director of network services and project director for Wireless
Andrew about the university’s groundbreaking work with mobile and wireless
technologies.
Syllabus: I’d like to speak
with you today about mobile and wireless technologies and their impact on the
higher education campus. Carnegie Mellon was among the first institutions to
install wireless networks. When did the university begin providing wireless
connectivity and how has it grown on your campus?
Charles R. Bartel: We actually
began with wireless in 1994, primarily as an experimental project for researchers
and a limited number of educators. We had received a grant from the National
Science Foundation to support a number of researchers who had gotten similar
grants for mobile and wearable computing. This was all pre-IEEE 802.11 radio
deployment. We had wireless in about five buildings on campus, and had deployed
about 75 access points for roughly 150 users.
Then in 1998, when the original IEEE 802.11 specification came out, we partnered
with Lucent Technologies, the provider of the pre-specification equipment. They
had learned a lot about deploying wide-scale wireless LANs and wanted to look
at how that would be done in the standards space. So, they granted us 400 access
points to refresh and expand in the academic campus. In 1998-2000, we deployed
a production wireless network in 30 buildings, covering almost 3 million square
feet with roughly 350 access points. And in 2001, we expanded the network into
our resident housing space, which added another 30 buildings and another million
square feet of interior space. So today, with the Wi-Fi network we call Wireless
Andrew, we have about 4 million square feet of interior space and we’ve
covered almost 100 percent of the campus—roughly 105 acres, both inside
and out, with 802.11b, at about 11 megabit speeds.
S: So Carnegie Mellon has ubiquitous
wireless coverage, then?
CRB: Yes, since Fall 2001.
S: How do faculty and students
use wireless today on your campus? Are they fairly mobile users of technology?
CRB: Yes. There’s a range
of uses, as you might expect, from more standard things, such as e-mail, Web
browsing, downloads, and instant messaging, to some of the more experimental
applications like proximity-aware and location-based capabilities, using handhelds
and PDAs. And we’re starting to experiment a little more with Tablet PCs.
S: What
were some of the original research applications and uses of wireless for the
mobility researchers?
CRB: There were projects with robots,
wearable computers, and experimentation with integrating eye pieces and using
different form factors of computers that can provide information.
S: Could you tell me a little bit
more about the wearable devices—what those are like, and what are some
potential applications?
CRB: There are many examples, some
of which you can see at www.wearablegroup.org. A part of our engineering school,
IECS (Institute for Complex Engineered Systems) looks at a number of applied
problems. For example, people that do any level of maintenance on very large
items—in the military you may be talking about a tank—may have to
refer to multiple books the size of a NYC telephone book with schematics and
blowout diagrams of all the parts. With a wearable computer, people can maneuver
around, physically, and still have access to a wealth of information through
downloads or the wireless connection to their wearable device. Another example
is our research into using a universal translator for speech-to-text or text-to-speech
translation of foreign languages.
S: Would the speech-to-text technology
have any application in the classroom, say in terms of note taking?
CRB: That is a potential, as what
you could be looking at is the ability to capture, in real time, what the lecturer
is saying, in addition to having access to what might be a PowerPoint slide
or a Web page of notes or things like that, and also being able to capture some
of the real time interaction between a student and a faculty member as they
are going through a problem or discussing a particular area. A project called
Informedia, led by Dr. Howard Watclar, has been pursuing research and development
in this area for many years at Carnegie Mellon. [see www.informedia.cs.cmu.edu/]
S: Is there anyone on campus doing
research into the pedagogical implications of mobile computing?
CRB: There is, particularly in
marrying the use of handheld devices with wireless communications capabilities
in the classroom. We have a professor by the name of David Yaron, who teaches
a modern chemistry class. For those of us who have taken chemistry, we know
that some learn and understand it pretty easily, and others, like myself, just
don’t get it as easily. Sometimes concepts need to be presented in a multitude
of ways in order to gain an insight. Prior to using handhelds, Yaron used to
pass out numbered index cards, and periodically take a poll during lectures
as to the level of understanding of the material just presented. If he saw a
vast majority of “1s” held up, he knew that the class got it. But,
there would still be some concern about the reliability of the poll—who
wants to be the one holding up a “4”? Now, using handhelds, the
poll can be answered somewhat anonymously, and tabulated as well.
The variant on that is that you can give ungraded pop quizzes to get at the
same information. Could you do all that in a wired classroom with a wired keypad?
Absolutely. But if you have a campus-wide wireless infrastructure and students
with handheld computers, you can set this up in any classroom, not just in the
few you have wired with keypads.
S: I’d think particularly
if the students are bringing their own PDAs, they could have their own profiles
and keep some data about themselves, too.
CRB: Absolutely.
S: What’s going on in terms
of location-based experiments?
CRB: Some of these experiments
have been done a couple of years back, but our purpose is to see if there are
some new and interesting ways of doing things. One example is that we had a
graduate engineering course that was working on a typical problem that students
face when working in collaborative teams. They may be in a number of different
teams with all different team members. And everyone probably has a different
schedule, so getting together to work may be a hard thing to schedule. But,
we can alleviate this somewhat with the use of handhelds and some collaborative
software. Let’s say that you have a project team. There is one student
who has been delayed. The team is able, through location-aware information,
to triangulate on the person’s whereabouts and instant message them via
the handheld, letting them know that the meeting is going ahead, and where it
is. And through some collaborative whiteboard software and the mobile student’s
and the team’s handhelds, they can actually conduct the meeting while
the late comer is in transit, and have a fair amount of that information beamed
to that person. They won’t have to restart the meeting or recap and have
that inefficiency occur. You might actually even ask that person to pick up
a remote printout or perhaps a resource at the library, while they are on their
way, then have them rejoin the team at some point. So that was one experiment
that was done in the engineering school. The various pieces of collaboration
software and the location-based triangulation and location-aware applications
were developed as part of the course.
S: Has there been anything done
with proximity-based technology in terms of campus information, where the location
itself can push out useful information to willing recipients?
CRB: Yes, actually there is a company
called Pango Networks, which is based in Pittsburgh, and we’ve piloted
some of their proximity-based or location-aware technology on our campus, with
our admissions staff. Many campus tours are done with present students acting
as tour guides, and while they may do an excellent job, a lot of the information
they convey is based on their own experience on the campus. The notion with
Pango was to build up a database of information pertinent to various schools
and buildings, so that when the tours walk around campus and approach a particular
building, they can get information about what departments are inside. They could
identify the faculty and their locations, and find out about their specializations.
All this is done as Hypertext links. The advantage is having the device understand
where you are by triangulation, on the wireless network, and beam you appropriate
information.
A variant of proximity-based technology that we’ve been doing some research
on is an application called Restaurant Concierge. Now this might seem a little
bit contrived, but all these experiments are made to test the feasibility of
the technology and how you might apply it further. Let’s say, if you are
familiar with the campus at all, you might know the location of a number of
eateries on campus. And your PDA probably has your calendar and schedule. What
Restaurant Concierge d'es is to take all of the information in your environment,
act as an agent, try to make some intelligent choices for you, and suggest things
for you. For example, it’s aware that it’s now winter in Pittsburgh—and
it may advise you that, while you have plenty of time on your calendar, you
might not want to walk halfway across campus, and there are a couple of eateries
within a five-minute walk of your present location. And BTW, here are the names
of two of your friends who are currently at one of the restaurants.
The advantage is having the
device understand where you are by triangulation,
on the wireless network, and beam you appropriate information.
S: What kinds of mobile devices
are most prominently in use by the general student population at Carnegie Mellon?
CRB: I would say that today, laptops
are the vast majority. PDAs are probably coming up in numbers, and Tablets are
probably in a third place at this point. But I think that still has to do with
people trying to decide what is the right form factor for them to use. So, most
students are going with a more traditional laptop, but as the power and flexibility
increases with handhelds, and as you are looking at Tablets maturing more and
more, I think that both of those are going to become viable competitors.
S: Which of the current mobile
devices that we see in some use today—laptops, PDAs, or Tablet PCs—do
you think will have the most acceptance or application in the future?
CRB: That’s hard to say.
The handhelds are very interesting from the point of view of offering a small,
light, increasingly powerful form factor, and have sort of an “instant
on” capability. In terms of writing, the Tablets show us more of a socially
accepted kind of behavior, allowing people to engage a little bit more than
a laptop might permit. It’s not quite clear how the human factors of these
mobile devices will eventually play out. At Carnegie Mellon, we are very concerned
to study the impact of the various devices on education rather than merely seeing
what develops. For this reason, we currently are starting a study on the use
of laptops in undergraduate architecture classes to see if ubiquitous laptop
access on a wireless network really had an educational impact or not.
S: What are the expectations of
Carnegie Mellon students for wireless connectivity? Do they come to the university
demanding ubiquitous wireless connectivity, including for residence halls, classrooms,
and public areas? Do you think that students in colleges and universities in
general in the U.S. will be expecting that kind of wireless connectivity at
their institutions soon?
CRB: At Carnegie Mellon, because
it is there, it is expected. It’s for the most part now an accepted utility.
When it is not there—a failure of wireless hardware, for example—we
hear about it. Now in terms of other colleges and universities, when I’ve
given talks and presentations with my peers one of the things that I tell them
is that wireless is going to show up on your campus. Either you as the campus
professionals are going to deploy it, or the students, faculty, and staff are
going to bring it and deploy it themselves. We are now looking at being able
to go to the local electronics store and for a couple hundred dollars being
able to buy an access point that we were buying a functional equivalent of for
$1,500 maybe four or five years ago. This technology has now become very cheap,
and wireless is also more common in homes. Students expect to see on campus
what is in their homes. If it is not in the university, people will start wondering
about the credibility of the institution. That may sound strong, but that’s
how it’s turning out.
S: In general, do you think that
mobile technologies will contribute to changes in models of instruction? For
example, might greater mobility foster different kinds of class meetings and
support collaborative work?
CRB: Yes, absolutely. With mobile
devices seamlessly connected to the Internet and the campus intranet, and as
people become more comfortable with the technology, they are going to find new
ways of presenting their material, and take advantage of the flexibility.
S: In the next few years, what
important trend might we see in wireless technologies in higher education environments?
CRB: It’s very hard to predict
the future. As wireless continues to grow, and we can put more information down
a wireless pipe, so to speak, I think we are going to see the applications become
a little bit more media rich. So whereas now we may be talking about downloading
Web pages, soon we may be able to do something more along the lines of animations
and streaming video.
S: What are the implications for
the convergence of telephony and wireless networks?
CRB: Telephony has some of its
own interesting requirements, as d'es video. Most data, particularly with IP
protocols, are more accepting of latency than audio or video. If the packet
d'esn’t get through, there is the provision for doing some retransmission.
Voice communications—like video—is not as forgiving in terms of
latency. So, what you are going to have to see for telephony to really take
off in the realm of wireless is more robust quality of service types of implementations.
And those are things that are being worked on by the IEEE now and will be eventually
incorporated into the existing wireless standards. So I think that you are going
to see VoIP, Internet telephony, or whatever you are going to call it occurring
in wireless technology. But you are probably going to be seeing that a little
later than on the wired side.
S: Carnegie Mellon is an urban
campus. What special opportunities d'es that present—with museums around
and different sorts of cultural resources you can connect to in the community?
CRB: There are real opportunities.
For example, adjacent to our campus there is a field left from an old baseball
stadium that was torn down maybe 30+ years ago. University of Pittsburgh and
the Carnegie Museum and Library are also in that vicinity. There is now some
amount of interest in making that a green park space, covered with wireless
to make it interesting for the community in general.
S: At Carnegie Mellon, you’ve
been consistently in the forefront in your wireless networking and in offering
the potential for mobile work environments to your students and faculty. How
will you extend this? What will you do next?
CRB: Because we are in the forefront,
the technology we deployed early is going to need a revamp. So we are looking
at the next generations of what are currently the ratified standards of 11a
and 11g. And we’re looking at how we can put in a composite wireless design
that can both serve the high-bandwidth needs, perhaps via an 11a infrastructure,
and still support what you might call the legacy users of b and g.
We’ll also be figuring out ways of taking advantage of the emerging area
of hot spots so that a campus community user can interconnect in a reasonably
secure and authenticated way back to the campus, riding on, as seamlessly as
they can, whatever radio LAN may be floating around where they are. There are
different models for that, and right now we are looking to find the best way
of integrating those, to meet future demands for seamless wireless connectivity.
We’ll keep looking into the best uses of mobile and wireless technologies.