Email Print Share


Photo of Arden Bement

Dr. Arden L. Bement, Jr.
National Science Foundation

Federation of American Societies for Experimental Biology (FASEB) Board Meeting
June 4, 2007

Greetings to everyone, and thank you so much for inviting me to speak with you. I truly appreciate FASEB's valuable work in promoting sound research and education policies.

When I thought about my remarks for today, I realized that I would follow a very lively discussion on large scale science -- and a very hungry crew. And sure enough, I was right! I see you are well on your way to satisfying your hunger, so I'll turn my full attention to the issues.

"Large scale" means different things in different contexts, and even in different fields of science and engineering. Let me say at the outset that the National Science Foundation values equally the contributions of single investigators and teams. We're careful to maintain a constructive balance between the two. Large scale science really transcends this distinction, since it increasingly accommodates both.

In my remarks today, I want to emphasize two developments that are driving large scale science: the continuing revolution in computer and communication science, and the growing convergence among disciplines in science and engineering. I'll conclude by drawing a few implications for crafting sound policy and budget decisions.

Large scale science is a topic of pivotal interest to the National Science Foundation. In fact, this sea change in the conduct of research is so critical to advancing the frontiers of knowledge that we have invested considerable intellectual capital to devise innovative ways to address the rapidly expanding scale of today's research enterprise.

That is altogether appropriate. Over the decades, NSF laid much of the groundwork for large scale science through investments in fundamental research across the entire spectrum of disciplines. But one field in particular has played a definitive role in facilitating large scale science. I refer to the revolutionary computer and communications capabilities that have transformed the contemporary landscape of research and education.

Thanks to these new tools, the discovery process now occurs at a tempo orders of magnitude faster than in the past. Only recently, researchers visualized the changing atomic structure of a virus by calculating how each of the virus' one million atoms interact with one another every femtosecond -- that is one-millionth-of-a-billionth of a second. It took about 100 days to generate just 50 nanoseconds of virus activity. Investigators say it would have taken the average desktop computer 35 years to come up with the results.

Just as important, simulation and visualization are powerful, new discovery concepts that open windows of understanding when experiment alone is insufficient or unfeasible. These methods help us build powerful predictive models of complex, dynamic phenomena -- from ecosystems to climate change, and from the cell to the human brain.

Early on, NSF funded the establishment of academic computer science departments to spur frontier research and train the talent necessary to expand this transformative field. Our support for the budding Internet as a communications tool linking scientists and engineers also drove wider collaboration, setting the stage for today's powerful interdisciplinary (and often international) teams, as well as distributed databanks and other tools.

We have come a long way since these salad days. Today, NSF supercomputer centers and the TeraGrid are at the forefront of scientific computing and frontier discovery. And the revolution is by no means complete.

In a sense, IT is a field in which success has continued to breed success. To put it another way, advances in IT have created the demand for yet more advanced IT -- and that is not a bad consequence. This is not merely a demand for more machines and sheer computer power, but rather a fundamental recognition of the intrinsic contributions to discovery and innovation that the computer and information fields continue to generate.

NSF has a number of irons in the fire to address these needs. We are now developing a national strategy for petascale computing to give scientists and engineers the resources needed to tackle their most computationally intensive research problems. Our GENI project, still in the early days, is the first step toward a next-generation Internet, with built-in security and advanced functionality.

In addition, NSF will soon begin the Cyber-enabled Discovery and Innovation (CDI) initiative to explore radically new concepts, approaches and tools at the intersection of computational and physical or biological worlds. CDI will address a wide array of challenges that include coping with massive amounts of data, exploring the potential of virtual environments, and educating an S&E workforce fully equipped to exploit the potential of computational discovery.

All of these projects taken together contribute to the development of cutting-edge cyberinfrastructure. "CI" -- as we call it -- helps scientists reach new frontiers in many ways, from support for modeling, simulation and complex data analysis to providing collaborative environments for geographically distributed teams of investigators. It simplifies the sharing of real-time and archived data, of customized analysis tools, and of the numerical models used in research.

And as cyberinfrastructure becomes increasingly pervasive and accessible, we can expect great transformations in education. Eventually, K-12 teachers, students or interested members of the general public will be able to participate in experiment and discovery in a way never before possible.

The broad and complex questions posed by 21st Century inquiry require access to these new information technology capabilities. They are critical for convergence among disciplines that is another defining feature of today's large scale, transformative research.

More and more, fundamental research at the interface among disciplines is proving to be the most fertile territory for discovery. Interdisciplinary research and collaboration are becoming the norm, rather than the exception, in many research endeavors.

The hybrid, interdisciplinary fields that emerged during the latter part of the 20th Century were harbingers of this fundamental shift. Bio-physics, bio-chemistry, bio-materials, bio-informatics, mathematical biology, computational biology and bio-geo-chemistry are now well-established fields. Neurophysics, neurochemistry and neuroeconomics are similar counterparts in the neurosciences. Several new and exotic entrants have just emerged, including synthetic biology, neuromorphic engineering, social cognitive neuroscience and nano-eco-toxicology.

Likewise, mathematics originating in the physical and social sciences, especially economics, is finding traction in a wide range of inquiries in the biosciences, across scales from the atomic to the global. Just consider the growing litany: statistical mechanics, quantum mechanics, information theory, game theory, graph theory, network theory, and advanced algorithms for computer modeling, simulation and visualization.

Brain science is one area in which this convergence is playing out. NSF recently held four workshops on new research frontiers in brain science at the interface among biological, physical and mathematical sciences, computer science, engineering, and cognitive science. For the first time, it is possible to envision comprehensive, multi-scale models of the brain that include relevant dynamics at different spatial and temporal scales. Investigators are exploring new methods that can link brain function to individual behavior and social phenomena, with extraordinary implications for learning.

This interaction is by no means a one-way street. In computer science, robotics and engineering, research is moving forward on a range of biomimetic devices modeled on brain function. NSF hopes to pursue the significant unexploited opportunities for mutual scientific benefit identified by workshop participants.

Of course, this to-and-fro among disciplines is not limited to brain science. We observe it in fields as diverse as molecular computing, and the design of novel materials at the nanoscale.

The prevalence of convergence and the power of cyberinfrastructure to facilitate discovery have definite policy implications. In today's research environment, we can't afford to neglect any avenue of discovery.

In addition to computer and information sciences, that certainly includes the physical sciences, engineering, and mathematics. Advances in all these fields are critical to next generation IT. The social sciences are just as vital. Investigators in many fields are now learning that human and social dynamics play out in the global arena with huge consequences -- from climate change, to emerging diseases, to issues of national security.

As you all know, funding for these fields has been stagnant for far too long. But that situation may be changing, with recent legislation that supports, among other measures, key aspects of the President's American Competitiveness Initiative. In particular, NSF is one of three agencies that funds research in the physical sciences, engineering and mathematics whose budget is slated for doubling over ten years. (The other two are DOE's Office of Science and NIST.)

I hope you agree that this is not a zero sum game. A balanced approach to assembling the nationís research portfolio keeps all boats afloat. Current proposals to bring more balance to that portfolio are promising precisely because they will move the research and education enterprise forward on all fronts. When that happens, everybody is a winner.

Scientific discovery and technological innovation often depend on the ripeness of ideas, as well as on new tools, and even serendipity. The past two decades have produced an unprecedented outpouring of new knowledge, with the promise of much more to come. The ITC revolution is still playing out, and the much anticipated nanotech revolution is sure to transform and disrupt the status quo in unimagined ways in the not too distant future.

But the productivity of the science and engineering enterprise also depends critically on less heroic, everyday factors. One of these is how NSF conducts its business of providing support to the community, and accountability to the public. If we fail to do this effectively, the community suffers, and so does the nation.

The simple truth is that the NSF workforce and infrastructure are beginning to show signs of wear and tear. Minor ups-and-downs aside, the NSF staff has barely increased over the past 25 years -- in fact, by only 4 percent. That is a mere 55 souls among some 1,300 in total. Meanwhile, the number of competitive proposals we process each year has increased by about 53 percent, to around 42,000.

NSF staff manages some 239,000 proposal reviews each year. These numbers are particularly noteworthy because proposals today present greater challenges. They address far more complex scientific questions, and involve a wider variety of collaborations from different scientific fields.

In recent years, NSF has made significant cost-cutting advances through the use of innovative information technology. NSF now receives and processes proposals electronically. Our IT systems now support over 250,000 users within 7,000 universities, colleges and other organizations.

We have managed to do all of this while spending less than 5 percent of our annual budget on administrative operations and awards management. Unfortunately, our ability to continue doing more with less is rapidly reaching the point of diminishing returns. We need to expand our workforce and modernize our aging IT infrastructure. We canít hope to serve the community or the nation at our existing high levels of performance without these human and IT resources.

Unfortunately, funding for NSF staff and infrastructure does not arouse passion in the S&E community or among decision-makers. That is understandable, but unwise. I hope you will agree that this funding is vital to the continued productivity of the communities you represent and that you will communicate this message to them.

Having treated you with a good dose of the practical, I'll conclude now with a leap to the visionary.

The nation's science and engineering enterprise has reached a tipping point. If we embrace the new paradigms for conducting research and education -- including large scale science, interdisciplinary collaboration and cutting-edge cyberinfrastructure -- we have every reason to believe that America will remain on the frontline of transformative research in the years ahead. And we can foresee progress in solving some of society's most persistent dilemmas. That will require unprecedented levels of interchange and cooperation across the entire spectrum of disciplines. I know I speak for all of us at NSF when I say that we welcome closer collaboration with FASEB communities in realizing this potential.

Thank you.