Major Research Instrumentation Program

FEBRUARY 22, 2005 SPEECH BY NSF DIRECTOR ARDEN BEMENT

Issues Related to the Discussion about an Interagency Program for Mid-Size Instrumentation

By Dr. Arden Bement, NSF Director

For the Committee on Advanced Research Instrumentation
National Academy of Sciences (COSEPUP)
February 22, 2005

I would like to thank Martha Krebs and the Committee on Advanced Research Instrumentation, for involving me in this ongoing discussion about the instrumentation needs of the research scientists and engineers of the Nation.

Before we have, what I hope will be a fruitful discussion about the issues at hand, I do have a few remarks I would like to make.

I must first note that the specific activity in which you are involved was requested by Congress in the NSF Reauthorization Act of 2002. And for context, this is the same Act that called for the NSF budget to double in a relatively short period of time.

• The Congress originally requested that the NSF “enter into an arrangement with the National Academy of Sciences to assess the need for an interagency program to establish and support fully equipped, state-of-the-art university-based centers for interdisciplinary research and advanced instrumentation.”

Your committee decided to broaden this charge and to focus the study on more current and contemporary issues, significant infrastructure issues facing our research communities today. In the discussion you have included not only interdisciplinary issues such as cyber-infrastructure, but you have also included the instrumentation needs of those within a discipline.

It is also worth noting that you decided to focus on research instrumentation greater than $2M and less than several tens of millions.

Therefore, your focus is on “Mid-sized instruments.”

Defining “mid-size” instruments may be something worth discussing. Should they be defined in terms of dollars or functionality? One might say that dollar limits are artificial and arbitrary. What may be mid-size in terms of dollars for one field of research may be small for another field.

Instruments are also treated differently in different environments. For example, a supercomputer at a major research university is treated differently from one located at an undergraduate institution.

So, defining mid-size does have its challenges.

The Committee may do well to put some other parameters around what you are designating as midsize. I might also suggest that databases, major survey instruments, and distributed or networked instruments not get lost in this discussion.

Regarding policy matters:

• NSF invests between 23-28% of its Research and Related funds into physical infrastructure; this has been the case for about the last 15 years.
• The National Science Board in its recent study, Science and Engineering Infrastructure for the 21st Century: the Role of the National Science Foundation, recommended we allocate funds to infrastructure more on the high side of this range. And, in FY 2006, we are requesting 27% of our R&RA funds go to physical infrastructure or TOOLS.
  • In our TOOLS category, we have the following:
    • Large facilities;
    • Infrastructure and instrumentation programs;
    • Federally Funded Research and Development Centers; and
    • Polar tools, facilities, and logistics.
    • This TOOLS category accounts for $1.53B of NSF dollars in the FY 06 budget request.
• We discuss the issue of portfolio balance annually with the National Science Board.

We feel this discussion is healthy, and a sound practice for deciding how to allocate funding for research and the infrastructure necessary for our investigators to carry out their work and to stay at the frontiers of science and engineering.

• You should be aware by now that NSF has about 15 different instrumentation programs that specifically support research instrumentation requested by the various communities.
• These publicized programs generally fall under the mid-size range in which you are interested.
• However, we do see within our instrumentation programs initiatives to support instrumentation in the size and range in which you are now focused.
• It is correct in stating that funds for mid-size instruments, as you defined them, are not easy to come by. These funds are currently in research program accounts. And program staff must make hard choices to start building instruments say in the $4M to $20M range. In order to initiate such action, it requires a ground swell from the specific research community.
• However, this is exactly what happens in most cases. The research community comes together and determines that a major investment is necessary in order to move the field forward, similarly to what occurs with instruments or facilities that fall within our MREFC account.
• Currently, we have no specific policy that focuses on mid-sized instruments as you have defined them.
• We do however insist on shared-use for major research instrumentation and management plans for instruments that are too expensive for single investigators.
• Our disciplinary program officers communicate extremely well with their counterparts at other agencies. Their relationships have led them to share the costs of large instrumentation with NIH, DOE, NASA, and NOAA, as needed, and even demanded by the community. These instrumentation projects are always selected using a merit review process.
• The National Science and Technology Council has engaged the issue of potential policy changes needed for the alignment of funding opportunities with scientific opportunities. And, after a Request for Information and 4 Town Meetings that provided input to the NSTC process, the following was ascertained regarding instrumentation:
  • 1) There continues to be an appetite for more modern instrumentation;
  • 2) From the faculty perspective, there still seems to be a need for instruments in the $35,000 to $500,000 range; and
  • 3) From both the faculty and administrators perspectives, obtaining operations, maintenance, and training support for major, complex instruments obtained through federal programs is a significant issue on college and university campuses.
• The Federal Demonstration Partnership at their recent meeting, focused on mid-sized instrumentation needs, and their main concern was the support for operations and maintenance, both over the short and long term.

[The issue of shorter life cycles for instruments also surfaced along with the additional complication of software upgrades associated with large, major instruments.]

Regarding cyberinfrastructure, this is a good example of how the infrastructure needs of our various research communities are expanding inside a non-expanding budget.

In FY 2006, NSF is placing a high priority on investments in cyberinfrastructure.

An infrastructure of power grids, telephone systems, roads, bridges and rail lines buttressed this nation’s industrial economy and allowed it to prosper. However, cyberinfrastructure- a networked system of distributed computer information and communication technology- is the lynchpin of today’s knowledge based economy. In FY 2006, NSF cyberinfrastructure investments total $509 million, an increase of 7.6% in a budget that is only increasing 2% over the FY 2005 level.

Modeling, simulation, visualization, data storage and communication are rapidly transforming all areas of research and education. NSF investments in cyberinfrastructure support a wide mix of projects and encourage participation from broad segments of the research community that rely on such technology as they tackle increasingly complex scientific questions. Thanks to cyberinfrastructure and information systems, today’s scientific tool kit includes distributed systems of hardware, software, databases and expertise that can be accessed in person or remotely. In fact, programs such as Teragrid, a multi-year effort to create the world’s largest distributed infrastructure for open scientific research, are specifically designed to transcend geographic boundaries and accelerate virtual collaborations.

The challenge facing the NSF regarding mid-size instrumentation is simple:

• As with all major instrumentation opportunities, there is a trade-off between allocating funds for the PEOPLE performing the work and the physical infrastructure critical for the progress of the whole field in question.
• Therefore, it is imperative that the specific scientific communities understand these trade-offs and are included in the discussion about large instrumentation projects.

As you well know, the various federal agencies have different missions. And those different missions have nuances that sometimes allow us to work well together and sometimes not. Any interagency program focusing on infrastructure should take this into consideration.

More funding is always desirable, but your report should not be about requesting more budget. It is clear however, that the fields that are being pushed for more complex infrastructure, more costly instrumentation that requires more operations and maintenance laden instruments are the same fields being stressed to find support for their workforce.

Based on my understanding of the problems and stresses regarding mid-size instruments, we should certainly be considering standards based on best practices for ensuring operations, maintenance, and staff training for federally supported research instruments (of varying sizes). This is necessary so that undue burden is not placed on our research community.

As you develop your recommendations, I hope you will consider the budget environment in which we find ourselves. Currently, it is apparent that any new program will come at the expense of existing activities. I have every confidence you will consider this during your deliberations.