Dr. Arati Prabhakar
IEEE Symposium on the NII
June 26, 1994
The National Information Infrastructure, or NII, is a huge enterprise, one that I could talk about for hours on end. Today, I want to sketch out the overall picture of the National Information Infrastructure, what the administration is trying to do in that area. That includes the work that I've been doing on the administration's Information Infrastructure Task Force. The other very closely linked subject is what my agency, NIST, is doing and planning in order to step up to the challenges of that information revolution.
Let me start with the big picture of the NII. These days, every time you open the paper there is an article about the information superhighway. Sometimes it's about the 500 channels of video. Sometimes it's about how people have gotten too absorbed in cyber space. What is clear is that suddenly there is a lot of discussion about the evolving NII. We need to realize that our society today already has in place an information infrastructure. We've had an information infrastructure for as far back as people have roamed the Earth -- perhaps starting with smoke signals. Fundamentally, the NII is the issue of how we communicate, how we share information. In fact, to be more correct, today we have several information infrastructures. There's one for broadcast, one for cable, one for our telephone system, one for newspapers, etc., etc., etc. And they each work after their own fashion.
So what's new about the NII today? What is different is that today we are imagining a future driven by the technologies that have been brewing and now exploding over the last few decades. We are imagining now a future in which we really have an information infrastructure that is a combination of all of those infrastructures that have been independent in the past. We are envisioning this new infrastructure that is transparent and interoperable -- where I, as a user, won't care about where the information came from or how it got to me or how it got translated onto my video screen. What I will care about is the content, the kind of information I am receiving. I will be able to find information as opposed to merely data. And that vision is different than the way technology and information infrastructures have evolved over the last many, many decades or even centuries.
Many people describe the infrastructure as telecommunications or how computing is changing. To me the NII is really the sum of all of those things. It's the sum of the ways in which we transmit, process, and share information.
If you think about it in terms of where the technologies are, one way to describe the information infrastructure is to imagine several layers. The first layer is bitways, all the ways that we have of moving bits around. That means wireless, cables, and fiber optics -- really the sum of all of those methods. That first layer also includes the networks and the protocols to make the networks function.
On top of that you might imagine a services layer which is made up of all the software tools that turn the bits into information. To me, that is in many regards the most ill-defined, but probably the most important, part of the entire information infrastructure. I think the simplistic view is to imagine the NII as being simply the physical networks. But a lot of the content, a lot of what will make it really useful, is in that middle services layer.
On top of those two technology layers you might imagine applications of the information infrastructure. In many regards that is the most exciting issue today: the fact that this is no longer just about technologists going off and creating more bits and talking about baud rates. All of a sudden this technology is beginning to touch people in very real ways. The usual areas of application that we talk about -- health care, manufacturing, electronic commerce, digital libraries, education -- that list is very extensive. But, if you sum all those different application arenas, the compelling message is that it is no longer about the technology in one or two areas. Suddenly we are at the brink of being able to see this technology pervade the way that we live, work, learn, and interact with each other. It is becoming the kind of capability that is fundamentally transforming our society.
The Clinton Administration is providing much of the leadership and the focus on this issue. It is a key priority. But I hope the message also has been very clear that we believe that it is the private sector in this country that is going to build, own, and operate the information infrastructure. The analogy to the physical infrastructure -- to the highway system -- in this country, is a useful one. But it's also quite limited in the details. This is actually a very different business, in part because today's version of the information infrastructure is largely private sector owned and driven. And we imagine that's what the future looks like too.
Having said that, why are all of us in government spending so much time talking and working on this issue? There are several reasons.
First are the various policy arenas in which the government is embedded, particularly telecommunications policy. We're in the middle of the first major reform in telecommunications policy in many decades. The government is in the telecommunications policy and regulation business because of a concern about universal service, a concern about access to these important new technologies. Sixty years ago it was a different set of new technologies. But it is the same idea, so today we worry about making sure that these new technologies don't create further separations in our society between the information haves and have-nots. Rather, we hope these technologies become a tool for greater coherence in our society. At the same time, we're trying to the greatest degree possible to limit the regulations in order to stimulate competition. There is a very strong feeling that the best way to make the technology move ahead aggressively, to make it cost effective and therefore pervasive and available to the largest part of our society, is to stimulate competition. With that in mind the administration is working very closely with Capitol Hill to try to get through one of the most dramatic changes in telecommunications policy that we have seen in a long time.
There are many other policy areas where the government has some involvement. Intellectual property is obviously an area that's very central to how we use these technologies. So are security and privacy issues.
The government also has other roles. We are a major user of information technology. If you think about what the federal government does with taxpayers' dollars for taxpayers, an enormous amount of our job is providing information, providing services. It's a very information-intensive business, so we are a major user of information technologies. And we can certainly be more aggressive and more coherent user of the evolving information infrastructure. That is one of the jobs the government needs to step up to more directly.
Finally, the government is a major participant in technology development in this country. The government provides almost half of the funding in this country for R&D of all kinds. In information technology the government has a major R&D role. And while we do this R&D in many instances to accomplish specific government missions, like national security, the bottom line is that it is a job that we need to do in very close coupling with the industry if it is to be effective at all. Each of these government roles in information -- setting policy, using information technology, and helping to develop new technologies -- all mean that the government is a player, a fairly important player, in the development of our national information infrastructure. Again, the context has to be very clear that we're a partner, that we want to be the best partner we can be to accelerate rather than drag that process. But it is surely not a job that we pretend to own by ourselves.
In addition to these concrete government roles, there has been one other very important development. Particularly with Vice President Gore's leadership and with Commerce Secretary Brown chairing the Information Infrastructure Task Force for the government, this administration is providing very senior attention on this issue. That has resulted in something rather unusual that was perhaps best captured for me by a colleague in industry. We were chatting with some people in his company and he said, "Gee, you know, the one thing that actually has happened with the NII is that the government has done the one thing that I could never imagine. They've created a vision." I go to meetings like this one involving all sorts of organizations. Of course, everyone knows that information technology is important. They have known that this revolution was coming. But today those conversations are more focused and there is a broader recognition that we need to be working across industry boundaries, that we need to be thinking about it at a national scale. In fact, we need to be thinking of it on an international scale. That conceptualization is very important, and it has been driven by this administration.
Most fundamentally what is different about what we are doing now is that everything we have done in information technology, or in any of these infrastructures that we have in the country today has been driven most often by a single company or at most by a single industry. What is different about this future-oriented vision is that it is by definition going to cross industrial boundaries. It redefines industrial boundaries. That is not the sort of thing that can be driven from one point. It needs to be considered on a national and international scale.
With that as a context of what the NII is and what we hope it becomes, let me give you a snapshot of the administration's specific NII activities. First of all, the very nature of what I have been describing is that it is extremely pervasive. I can't find a part of the federal government that is not involved in one way or another, at a minimum as a user of the technology. Recognizing that this is not something that any one agency is just going to go off and do by itself, the administration created the Information Infrastructure Task Force, or IITF, that Commerce Secretary Brown chairs. This is the focal point for all the players across the government that have any major NII involvement, and it is a focal point that just did not exist previously.
The IITF is organized into three committees. The Telecommunications Policy Committee is chaired by Larry Irving of the National Telecommunications and Information Administration, the Commerce Department agency that is at the heart of all the telecommunications reform efforts. Sally Katzen, of the Office of Management and Budget, chairs the Committee on Information Policy that is concerned with intellectual property, privacy, security, and related issues.
I chair the third committee, the Committee on Applications and Technology. Our job is to coordinate the huge range of application activities across the federal government, which, of course, necessitates the complementary role of coordinating discussion of policies that underpin this vast assortment of technology activities.
For starters, we asked ourselves, "How do we get our arms around all these applications?" One option was to publish a catalogue of all government applications of the NII -- both internal agency applications and demonstration projects with the private sector or other public organizations. We quickly realized that the minute you print a document that contains all those application activities, it is instantly obsolete because everything is changing so fast. So, we did something novel. We used information technology and created an on-line inventory, accessible to people inside and outside the government. If you hunt through this inventory, you'll find plans and activities that range from a twinkle in someone's eye to things that are actually up and running and real today. You'll find activities in every possible application arena, from health care to education to manufacturing.
The Committee on Applications and Technology -- CAT, for short -- also has issued several documents. A very early report examined key barriers to creating applications in specific arenas. Not surprisingly, many issues were raised: is the technology ready, are people ready to use it, and so on. To me, one surprise was the focus on the adoption of technology -- on how people use the technology, on how technology can be tailored to ensure ease of use, on how organizations learn to use technology. This focus reflects growing recognition that you can't just plop a bunch of hardware and software in and assume that good things will begin to happen immediately.
In May, CAT published a second report, called Putting the Information Infrastructure to Work. It's a collection of seven white papers written by experts in the government that are working in various application arenas. Each paper articulates the benefits made possible by an advanced information infrastructure. In addition, for each of the seven application areas, the report summarizes current activities that can help make this vision a reality. What's happening today? What projects are under way? What still needs to happen in the near term, and longer term? Then each paper poses questions to readers, asking for views and ideas on government policies, programs, and investments that will enable the development of the NII and foster its best and fullest use. As intended, the document is stimulating a much broader dialogue with the private sector and with the general public.
CAT also has a set of working groups, which is where the action is. One of those is the government information technology services working group, headed by Jim Flyzik, the Treasury Department's director of telecommunications management. Their admirable goal is to coordinate and improve how government agencies use information technology so that, for example, members of the public can easily negotiate their way around government information kiosks and go directly to the service they are seeking, without having to master a befuddling array of agency- specific formats.
The CAT working group on health information and applications is trying to corral the various agencies spread across the government that have a role in applying the NII to health care. John Silva, program manager at the Advanced Research Projects Agency, is the chair.
A third group -- the technology policy working group -- is a neutral forum within the government, where agencies can focus on issues bearing on the interoperability and scalability of new information services. These are absolutely critical technical issues for the NII vision, and the working group is engaging the private sector in the search for solutions. Questions abound, and finding answers requires looking at the whole concept -- not at a particular point in technology.
The technology policy working group is chaired by Duane Adams, who is the ARPA deputy director. It has initiated four projects. A big-picture effort -- one that must, and does, involve industry -- aims to create an NII roadmap. If all these technologies are to miraculously merge and work together, it's useful to have some sort of a top-level roadmap that says when people expect certain capabilities to occur. Another project focuses on issues affecting how video technology connects to the infrastructure. A third activity concentrates on services architecture, situated in the middle layer of my three-layer NII cake. That's where I think a lot of the very challenging issues are. And I also think that not enough conceptual work has been done to define how we create true scalability and interoperability in that part of the system.
Finally, there is a project that has begun to tackle the issue of standards -- a tremendously important area. I have not been to a single NII meeting in the last year that failed to mention standards. People are not talking about standards because they love them and they are working; they are talking about them because they do not work, or at least not work nearly well enough. At NIST, we know well that if standards work people don't talk about them. So, the widespread concern over standards is a real flag, a sign of serious need. Remember that we are really trying to accomplish something unprecedented in terms of interoperability. We are talking about an enormous system -- one of a scale and scope never tackled before. So, looking at standards and the standards process more fundamentally is also something that the technology policy working group is engaged in, and NIST's Jim Burrows is leading that part of the activity.
By now, the enormous dimensions of the NII should be clear. The challenges are huge, commensurate with the tremendous benefits that can be reaped. To clear the obstacles and to make lasting progress, you must first zoom your lens back and get a systemwide perspective. Then, you can really begin to appreciate why no single player--company, industry, or government--can make the NII happen. If we do not establish an active dialogue that involves players across many, many domains, then the sum of our fragmented individual efforts could easily be overwhelmed by the sheer magnitude of the obstacles.
At NIST, we intend to be a handmaiden to this cooperative process. Why? For two reasons. First, because that's our job-- to work with industry in developing and applying technology, standards, and measurements. Each of those elements is a vital component of efforts to realize the NII vision. Second, an advanced information infrastructure would help us do our jobs better and, I think, with greater effect.
Take NIST's Manufacturing Extension Partnership, for example. The MEP is designed to be a comprehensive, yet locally responsive, network that helps smaller manufacturers upgrade their equipment, improve their processes, and strengthen their business performance. The local, physical nodes in this network are manufacturing extension centers, numbering about 35 now and 100 by 1997. One of the long-term dreams in manufacturing extension is to couple the nation's 370,000 small and medium- sized manufacturers to the information revolution. Think of the opportunities that would create for companies and their customers, and for American competitiveness in manufacturing.
The MEP has a well-thought-out information infrastructure strategy, and it has launched a number of pilot and demonstration projects that, I think, will pave much of the way to the electronically integrated manufacturing extension network now envisioned.
At the same time, NIST's Advanced Technology Program is helping the private sector develop some of the high-risk, enabling technologies that industry believes will help it seize the opportunities created by the evolution of the information infrastructure. The ATP provides cost-shared funding for competitively selected technology development projects proposed by individual companies or joint ventures.
In its first five years, ATP functioned as a pilot program, using its modest budget to help support projects chosen from merit- based general competitions open to any and all areas of technology. The President aims to increase the program budget to $750 million by 1997. This year the budget is $200 million, up from $68 million in 1993. The President has requested $451 million for 1995.
These actual and anticipated increases are enabling ATP to focus its benefits for maximum impact. Starting this year, the bulk of funding will be invested in multi-year, multi-project efforts focused on achieving specific, well-defined technology and business goals. We will continue to hold at least one general competition each year, but we believe that by funding and managing groups of projects that complement and reinforce each other, the ATP can have a greater impact on technology and the economy.
We inaugurated this strategy earlier this year, announcing a set of five focused program areas in which the ATP will begin investing. Three of the five pertain directly to matters of information infrastructure. One is focused on helping to trigger an information infrastructure for health care, where information processing is conservatively estimated to account for 20 percent of total costs. Another aims to leverage and extend information technology applications to help industry achieve truly computer- integrated manufacturing of electronics. The third program--the ATP Component-Based Software Program--aims to foster development of technologies needed to enable systematically reusable software components suitable for a broad array of applications. Today, about 85 percent of software was developed from the "ground up" for costly, customized applications. Progress catalyzed by the ATP software program, I think, will be critical to realizing the NII vision. The other two programs--health care and CIM for electronics--will help drive the capability in these very important application areas.
Let me turn to what's happening in the NIST laboratories. NIST has a 93-year tradition of working with industry, a tradition built on the high-quality physical science and engineering work carried out in its labs. The greatest value-added generated by the NIST labs is in the form of infrastructural technologies--the underpinning technologies that raise the level of play for everyone. This job is critical. Our greatest successes are the measurement techniques, the test methods, and other laboratory products that industry comes to take for granted because they are embedded into each company's own infrastructure. Yet, we know that this is precisely the kind of technology that goes begging for private investment because the returns are widely shared, unable to be captured by the innovator.
In the information technology area, I can point to some successes, with significant dividends reaped by companies: large firms like Intel, Hewlett-Packard, and Corning, and smaller firms like Optical ETC, and Optex and Analogy. Clearly, there is much more that our laboratories can, and must, do in this area, building on our work to date. For example NIST's MULTIKRON chip is being used by developers of massively parallel computers to help them identify processing latencies and bottlenecks. We are providing benchmarking tools for companies working to improve the spoken-language capabilities of computers, as well as their ability to recognize speech and handwriting.
And each month, some 20,000 outside users traverse the Internet to use NIST's Guide to Available Mathematical Software.
NIST's involvement in the development of ISDN (for Integrated Services Digital Network) illustrates another role that the laboratory program plays. In 1988, NIST formed the North American ISDN Users Forum. By that time, ISDN had already been rumbling around for quite awhile. It had become painfully clear that vendors and users in many very different areas needed to work together if broadband ISDN was really going to come into being as a national service. But they weren't. Sound familiar?
The first step NIST took was to convene everyone at the table--a neutral site. We brought together vendors from many many different areas, computer manufacturers, network people, and so on. That really hadn't happened before. At the same time, we brought in the user community. People from financial services, health care, all the different places that could really use this advanced capability that was brewing in ISDN. That was the beginning.
What has come out of that is a coordinated set of activities, many of which are carried out by the participating companies. NIST also did some of the important technical work that carried the North American ISDN User's Forum forward -- work in areas like conformance testing and application profiles. Both contributions -- convening the users and vendors and providing the technical underpinnings -- helped to catalyze the entire activity. ISDN continues to move towards wide-scale implementation. It's been a long and slow process in many regards. But I think that the North American ISDN User's Forum really did help to end the stalemate and to broaden the availability of ISDN and its applications.
Today, the need for this kind of activity is great, in both the technical and standards-setting arenas. That means we at NIST have our work cut out for us in three major program areas--the Advanced Technology Program, the Manufacturing Extension Partnership, and our laboratory R&D program. In our labs, we have to focus on delivering more value to information technology industry. We are stepping up to the challenge of providing the capability-building infrastructural technologies that this vast and growing industry needs.
The diverse information-technology industry is very R&D intensive. It surely is not the job of any of our programs to offset that private sector investment. That would be A) inefficient and B) inappropriate. But there are many areas where no one company can invest sensibly--because, for example, the returns are broadly shared or because of the magnitude of the task and the high-degree of risk it entails. We must always ask ourselves, first, if and why we need tax payers' dollars to do this job. Second, we must also ask whether our agency and, in particular, this NIST program is the best place to get it done. Among the plethora of tasks to be accomplished and the series of obstacles to be overcome on the way to realizing the NII vision, I think there are many jobs that pass both of these tests. And, in partnership with industry, NIST is eager to take them on.