Tom Finholt: OK, thanks Dan. I want to talk with you today about an experiment that's been going on here at the University of Michigan for the last five years. It's really a discussion about a new way of doing work, the collaboratory.

The question that we ask in this project is really, are we entering into a new era of organizing collaboration? I think we've heard today both through Vint Cerf's talk and through Doug Van Houweling's talk and through the digital library demonstration that the Web is providing tremendous new resources. The question that we might ask ourselves is, what gains are there for knowledge workers? They're possible, but they're not necessarily automatic. And the great challenge that confronts us here at the University of Michigan is, how do we maximize the likelihood of positive change with the introduction of these technologies?

Now before we try to address some of these questions, let me share with you one of my favorite pictures. This is an illustration, a depiction of the world, and what you see on the Y axis here is a scale representing the capability of information technology, while the X axis represents time.

That first curve is something I call the Wired magazine curve; it represents hype, or the kind of wild speculation that we often hear about information technology.

The middle curve represents performance potential, such as the number of instructions per second and so forth. That bottom line represents the realized performance that we experience on our desk. This is the minutes of cursing at Microsoft Word when it dumps our document; it's waiting for the Web; and so forth. And you'll notice that gap between the realized performance and the potential performance is something I'm labeling "the reality gap."

We think at the University of Michigan we're on to a methodology for addressing that gap, and that is to say that it requires a kind of interdisciplinary synergy that brings together expertise from user domains, expertise from behavioral science, such as myself - I'm a psychologist - and expertise from the domain of computer and communications science. In a moment, you'll hear from my colleague, Farnam Jahanian, on that score.

These complementary domains are brought together for us through the collaboratory idea, and Dan Atkins has already shared with you this picture. The idea here is that we use media-rich technology over the Internet to bring people together with other people, to bring those people together with the resources they need that manipulate the physical world and measure the physical world, and also those information resources, such as digital libraries and electronic documents that Wendy Lougee has just talked about.

This is brought together for us and organized through a project focused on space science. In particular, space science is an excellent test bed for this activity, because of a transformation occurring in this particular discipline. They are at a crossroads; they have been typified for the last century as a science of empirical observation, moving now toward a period of predictive capability. And just as a reflection of the importance of that, there are almost a trillion dollars worth of satellite resources in geosynchronous middle- and low-Earth orbit, vulnerable to effects of the solar-terrestrial system, and we have some indication of the degree of that vulnerability - those of us who have pagers lost them recently when the Galaxy 4 satellite went out. Imagine if we lost the entire global communication grid; this would be catastrophic.

In this case, to build a predictive capacity, to be able to anticipate changes in the solar-terrestrial system, requires bringing together a global array of instruments and globally distributed scientists, so that's the goal of this enterprise.

Let me share with you a little bit of the nature of the solar-terrestrial system, and this will be a brief digression. For those of you who slept through high school physics I'll try to keep this at a level that's interesting. I think it also reflects the truly interdisciplinary integration in this project in that I went from knowing nothing about space physics to where I think they could now at least confer a master's degree on me.

This is a diagram of the solar-terrestrial system, and all you really need to understand for purposes of understanding the collaboratory is that the Earth is surrounded by belts of magnetic forces. These forces capture particles that are being sent out by our sun, and the sun as represented in this diagram is a very dynamic creature. It is a star that's very important to us, and it goes through cycles of change. And what you see in the photograph here, dated September 1991, is an X-ray image of the sun at solar maximum. So this is a period of high solar activity, and what you see contrasting in the other picture is the sun at a period of solar minimum, or low activity. And these changes in the dynamic behavior of the sun have tremendous implications on Earth. In a simple sense, we experience them as the Northern Lights. Here's a very nice picture of the aurora taken from the ground in Alaska, and here's a somewhat rarer view, taken from the Space Shuttle.

So, in a nutshell, what do space scientists do? They study the aurora, the shape of these Northern Lights, the transformation and effect of the sun's energy on the Earth. The collaboratory, then, becomes an excellent tool for them.

How did these people used to do their work in the past? They would have abandoned their families, friends, colleagues and warm weather for weeks of research at cold, isolated and remote locations. This is a picture of the observatory in Kangerlussuaq, Greenland, which is a few degrees above the Arctic Circle, not a very pleasant prospect and also not one that lends itself nicely to convenient collaboration.

What they have today is this, or interaction through the collaboratory interface. Let me walk you through some of the critical features here.
What you see in the upper corner is something we call "the session manager." This is a representation of virtual rooms or salons, if you will, where the scientific activity is taking place. What you see in these images is the first-ever, simultaneous, real-time display of the entire Northern Hemisphere radar observatory chain, and what you see down in the lower corner is perhaps the most interesting part of this interface, the discussion that's going on 24 hours a day amongst this international community of scientists. So it's hard to convey from this still image the excitement that this provides for these investigators. It's really a new way for them to do their science over the Internet.

Let's talk a little bit about some of the impact here. I want to share with you two particular kinds of observations. The first one is that this tool, as Doug van Houweling and Vint Cerf have already alluded to with respect to the Internet in general, is an incredible mechanism for generating new kinds of links and access to other people, in this case to other scientists.

In the old days, this kind of access would have been limited by time and distance, and you frequently saw research that was typified by single scientists doing solo experiments, and among students it was a watch-but-don't-touch orientation. Today, through the introduction of this collaboratory, you have the capability of accessing diverse expertise in real time during ongoing experiments, and from the point of view of students, you have an incredible opportunity to access the leading scholars and scientists in your field while they're engaged in authentic research activity. Just to toss out a high-jargon term here, we might describe that as "legitimate peripheral participation," or for those of you who aren't conversant with that, you might describe this as apprenticeship. This collaboratory is a world-class opportunity for these students, often even in the first week of their graduate education, and in some cases as undergraduates, to experience research with people that they've read about in textbooks or seen give lectures and so forth.

Here's an illustration of what I mean. This is a pattern of communication that occurred in a recent experiment. What you see here are lines representing the frequency of interaction between sites. This gives a good sense of the global scope of this activity. To put some names and faces on those dots, the one there in Colorado is a graduate student, and he is interacting with colleagues in Massachusetts, in Norway, in Texas and on the West Coast, over the course of about eight or nine hours during a coronal mass ejection in April of '97, so a very priceless opportunity for him.

I want to share with you one last impact, which is to say that the collaboratory has, I think, forced a change in the experimental paradigm in this field. In the past, space science would have been typified by what I describe as a pinhole camera orientation: you're very fixed on your instrument, with the consequence that your view of the atmosphere and of the universe is strongly bounded by what your instrument can show you. But I think what's happening with the development of the collaboratory is that these scientists are starting to expand their perspective to adopt a truly global orientation, and this is shifting their emphasis from single instruments and what single instruments tell to the story that a chain of instruments tell, acting, if you will, as a collective kind of instrument.

Just to illustrate that, real briefly here, this again is a snapshot taken from actual use within the collaboratory. What you see on the top is a real-time image of an ultraviolet imager from a polar orbiting spacecraft, in this case operated by NASA. What you see in the bottom is a time-synchronized view of a supercomputer model of the same view of the atmosphere. So what scientists can do with this capability is compare their predicted data with the observed data, and then in some situations take that observed data back in and refine the model. So this is a tremendous capability.

But it doesn't come cheap, and the underlying technology is very sophisticated -- which is a good lead into my colleague, Farnam Jahanian, who's going to tell you about the nuts and bolts that underlie some of the complex data transfers and interactions that make the system possible. Farnam.