Special Report - Part 2: O.R. & the National Science Foundation

More on opportunities for research funding, current trends and new initiatives.

NSF

By Michael C. Fu and Russell R. Barton

Editor’s note:
This is the second in a two-part series on the NSF and relevant programs and funding opportunities for operations research.

Perhaps the most frequently asked questions we receive have to do with where does a particular proposal fit, i.e., where should it be submitted in NSF. Among the three programs that we cover, in a nutshell we differentiate between the OR and SES/MES programs as follows:

The OR program supports research concerned primarily with fundamental methodologies potentially motivated by applications, whereas the SES/MES programs seek actual domain-specific applications where the methodology and modeling are linked to characteristics of the application. Projects in the OR program may be hammers looking for nails, as long as the hammer is being well developed. Rather than having a specific hammer looking for a nail, projects in the SES/MES programs seek to find custom hammers that address a specific problem (of broad interest) in manufacturing or services.

For some projects there is not a clear dichotomy. Sometimes a project can propose to do both. In this case, the proposal might be reviewed by both programs and possibly co-funded by several programs. The tables of samples of currently funded projects include an indication of several projects that are co-funded.

Besides the CAREER and unsolicited proposals, which comprise the “bread and butter” of the core programs, we summarize other funding opportunities of potential interest to the OR/MS community:

  • Broadening Participation Research Initiation Grants in Engineering (BRIGE) program – This is another program for junior faculty that is less known, as it has a lower level of funding and has some particular requirements implied by the name of the program. It generally has a deadline around mid-January.
  • Research Experience for Undergraduates (REU): This is an ideal way to obtain supplemental funding attached to an existing NSF-funded project to support undergraduate students pursuing research, so if you find a good student in your undergraduate class, this is a marvelous way to introduce them to wonderful world of research. Again, contact your program director for guidance if you’ve never submitted an REU request before.
  • Early Grants for Exploratory Research (EAGER): This is a funding mechanism “used to support exploratory work in its early stages on untested, but potentially transformative, research ideas or approaches. This work may be considered especially “high risk-high payoff” in the sense that it, for example, involves radically different approaches, applies new expertise or engages novel disciplinary or interdisciplinary perspectives. These exploratory proposals may also be submitted directly to an NSF program, but the EAGER mechanism should not be used for projects that are appropriate for submission as “regular” (i.e., non-EAGER) NSF proposals.
  • Rapid Response Research (RAPID): This is a funding mechanism “for proposals having a severe urgency with regard to availability of, or access to, data, facilities or specialized equipment, including quick-response research on natural or anthropogenic disasters and similar unanticipated events.” Examples include research related to the aftermath of Hurricane Katrina and the Japan earthquake and tsunami.
  • Grant Opportunities for Academic Liaisons with Industry (GOALI): This is perhaps the best opportunity to leverage research that involves collaboration with industrial partners. In the OR, MES and SES programs, GOALI proposals go through the same panel review process as that of unsolicited proposals.
  • Industry/University Cooperative Research Centers (I/UCRC): This program offers another much larger in scope opportunity to involve collaboration between academia and practitioners.
  • CREATIV: This is a brand new initiative that was launched in fall 2011 as the first part of the INSPIRE initiative by the NSF Director Subra Suresh. Please refer to the FAQ page for CREATIV. Note that a formal mechanism is in place for alerting program directors of your interest and intent for submitting a proposal under this program; see http://www.nsf.gov/od/oia/creativ/inquiry.cfm.

GOALI and I/UCRC were discussed in the Klein/Smith August 2009 article, and current information can be found on the NSF Web site.

Multi-disciplinary programs offer another avenue for funding OR/MS research, as well as an opportunity to work with researchers in other fields. A sampling of some relevant programs include the following:

  • Cyber-enabled Discovery and Innovation (CDI): Although this program no longer takes any new proposals, having awarded its last funding last year, it supported “interdisciplinary computational and data-enabled science and engineering.”
  • Innovation Emerging Frontiers in Research and Innovation (EFRI): “Embedded Distributed Simulation for Transportation System Management,” “Theory and Algorithms for Autonomous Reconfigurability of the National Air Transportation System.” The 2012 topics are Flexible Bioelectronics Systems, Origami Design for Integration of Self-Assembling Systems for Engineering Innovation and Photosynthetic Biorefineries.
  • Science, Engineering and Education for Sustainability (SEES): This is a new NSF-wide initiative with the following mission statement: “To advance science, engineering and education to inform the societal actions needed for environmental and economic sustainability and sustainable human well-being.” For example, Energy Efficiency & Management is one of the topic areas listed in the new program, Sustainable Energy Pathways.
  • Cyber-Physical Systems (CPS): Cyber-physical systems “are engineered systems that are built from and depend upon the synergy of computational and physical components… The goal of the CPS program is to develop the core system science needed to engineer complex cyber-physical systems upon which people can depend with high confidence.”
  • Cyberinfrastructure Framework for 21st Century Science and Engineering (CIF21): This is an NSF-wide crosscutting portfolio of activities just getting underway, with a vision to “provide a comprehensive, integrated, sustainable, and secure cyberinfrastructure (CI) to accelerate research and education and new functional capabilities in computational and data-intensive science and engineering, thereby transforming our ability to effectively address and solve the many complex problems facing science and society.” It is viewed as a successor to the CDI program.

NSF also supports complementary research activities such as workshops and conferences and travel grants. Please contact the program directors for guidance before submitting any proposals in these areas.

Opportunities, Trends and Initiatives

Some of the research trends that the programs are emphasizing include new funding initiatives in healthcare systems and recent (or upcoming) workshops.

Operations Research: In addition to continuing strong support of the traditional twin pillars of optimization/math programming and stochastic modeling, one current trend of the OR program is the to encourage research that better integrates the two perspectives. One such direction falls under the broad umbrella of dynamic decision-making under uncertainty (see table of currently funded projects), which includes research that may still avoid stochastic models a la robust optimization. Two workshops along these lines include one on simulation optimization held in May 2010 (http://users.iems.northwestern.edu/~nelsonb/workshop/) with a follow-up meeting in December of 2010 and a forthcoming workshop May 31-June 1, 2012, entitled, “A Conversation between Computer Science and Operations Research on Stochastic Optimization.” This workshop is co-funded and co-sponsored by the Robust Intelligence (the new name for Artificial Intelligence (AI) at NSF) Program in CISE, and is motivated by a mutual feeling by the Robust Intelligence Program Director Sven Koenig and the OR Program director that the AI and OR communities carry out research on essentially the same topics oftentimes unaware of the contributions from the other community due to differences in terminology and notation. Such differences prevent research ideas from being understood and shared. This workshop aims to break down such barriers, focusing on one broad area of interest to both communities, encompassing well-known topics such as AI planning and Markov decision processes.

Service Enterprise Systems: The recent emphasis on healthcare systems engineering has many drivers, including recommendations from the President’s Council of Advisors on Science and Technology and the National Academies. OR/MS researchers will have new opportunities for funding if they collaborate with researchers in health information technology, health services, health operations, health policy, social and behavioral sciences and medicine. Renewable energy delivery and management remains an important area, as does work to understand the role of the Web in market characterization and service delivery. The Web also plays a role in “Enterprise Situational Awareness” – that is, a real-time understanding of the state of all aspects of a service operation: status of systems, demand volume and patterns. The opportunity for OR/MS researchers is to use analytics to construct a parallel, perfect-fidelity model of the enterprise, updated in real time, that allows what-if analyses for decision support in the face of unforeseen events such as equipment breakdown, supplier problems or demand shifts. Of course, research on enterprise situational awareness applies in the manufacturing setting as well.

Manufacturing Enterprise Systems: The G8 has an initiative on material efficiency as a first step toward sustainable manufacturing, and NSF is participating in this initiative. How to capture sustainability issues in OR/MS methods for manufacturing enterprise design and management will remain an important research area. A related area is the design and operation of energy manufacturing enterprises, for example biofuel production and solar and wind farms. Nanomanufacturing process management and control remains an important area for NSF and for the MES program. A closely related area is biomanufacturing of tissues and other biological products. As for SES, it will be increasingly important for OR/MS researchers to collaborate with other disciplines. For example, NSF’s Chemical, Biological, Environmental and Transport Engineering Directorate and the American Institute of Chemical Engineers co-sponsored a recent workshop on the design of sustainable product systems and supply chains, where the role of OR/MS was recognized.

Writing a Proposal

It is critical that your proposal be considered in the correct program, so that the appropriate experts and peers can review the intellectual merit and broader impacts of the proposed work. Along these lines, it is advised to go to the Web site, www.nsf.gov, to: 1) read the program descriptions, and 2) do a search of recent awards in the program by choosing “Search Awards” under the Awards tab, select the Program Information tab, and selecting the programs (separately) for which you believe your proposal might fit into the scope. In addition to the three programs in CMMI described here, consider the other programs mentioned earlier, as well.

If you are not sure where a proposal idea might fit, use the search awards option at the top right of the NSF page (it is a drop-down choice). In the search box, type a phrase that is related to your research idea, like “sustainable manufacturing” or “smart grid” – be sure to include the quotes.

If you have questions, don’t hesitate to contact the program directors by e-mail or phone. It is usually best to forward a 1-2 page description of your idea before trying to call.

Of course, finding the right program is not enough to be successful. A necessary but not sufficient condition for getting funded is having good research ideas. The other key ingredient is writing a good proposal. Some specific tips given in the August 2009 OR/MS Today article by Cerry Klein and Bob Smith still hold today, so those will not be repeated here. Writing a good project summary can make the difference between your proposal being given careful consideration or being dismissed early in the “triage” by a panelist. The Heilmeier Catechism gives very good guidance for writing the project description. You can find it via Google, but the basic points are as follows (slightly modified):

  1. What are you trying to do? Articulate your research objectives using absolutely no jargon.
  2. How is it done today, and what are the limitations of current practice?
  3. What’s new in your approach and why do you think it will be successful?
  4. Who cares? If you’re successful, what difference will it make?
  5. What are the risks and payoffs?
  6. What is your game plan? Tasks, sequencing, timeline? Contingency plans? Make it clear that the game plan will enable you to achieve your objectives.
  7. How much will it cost? How long will it take?
  8. What are the midterm and final “exams” to check for success?

In our experience, failure to address one or more of points 2-4 is the most common flaw in unsuccessful proposals.

As you write your proposal, make sure to refer to the Grant Proposal Guide for specific details. For the core programs, there is a 15-page limit on the project description, and there are font, spacing and margin requirements to be followed. Clearly indicated descriptions of intellectual merit and broader impacts, both in the project summary and project description, are required. Many proposers do not understand what NSF means by broader impacts: impact on society, outreach, education (graduate, undergraduate, professional and K-12), and the involvement of underrepresented minorities and women in research. Two new requirements that have been added recently are the Data Management Plan and the Postdoctoral Mentoring Plan, if postdocs are part of the proposed project. Due to the limited budgets, generally the three core programs (OR, MES, SES) do not encourage support for postdoctoral fellows from the unsolicited proposal funding. Program directors and reviewers like to see doctoral student support, and this is a critical part of broader impacts, with its emphasis on educational and societal impacts.

In CMMI, proposals that do not meet the criteria will be returned without review.

Brief Snapshot of the Panel Review Process

After the proposals have been compliance checked by the NSF staff, the program directors look at them to decide whether or not they are best served by being reviewed in their program. If they believe that another program is a better fit, they may contact the other program director and if that program director agrees with the assessment, transfer the proposal and inform the PI of the action. Alternatively, a proposal may be “shared” with another program, which would allow it to be reviewed in both programs (two panels) or just reviewed in one but possibly co-funded by the other program.

After the best home for each proposal is settled, the program director groups the proposals according to some criteria. For CAREER awards, the three programs considered here (OR, MES, SES) generally have a single panel for each set of submissions. Unsolicited proposals are generally grouped according to methodology or application. In the case of OR, this is usually done in two groups corresponding roughly to optimization/math programming and stochastic modeling, where some proposals (e.g., global optimization using randomized methods or approximate dynamic programming) could fall into either category. In the case of MES and SES, the panels might be something such as healthcare, energy, finance and supply chain management.

Once the grouping is done, the program director decides on panel dates and then invites panelists to serve. Once the panel is formed, the project summaries of the proposals to be considered in that particular panel are sent to the panelists, who then provide a preference for reviewing each of the proposals, upon which is based the proposal review assignments. Typically, a panelist will review between six and 10 proposals for a panel. The panelists then read their assigned proposals (and they are allowed to read any of the proposals in their panel) and assign ratings (excellent, very good, good, fair, poor) and a written evaluation of strengths and weaknesses along the dimensions of intellectual merit and broader impacts, concluding with a summary statement providing their overall assessment as reflected in the rating given. These are to be done prior to the panel meeting date, preferably at least three days in advance to allow the program director time to read and digest the various evaluations in preparation for the panel meeting.

For the three programs described here, the panel meetings are generally one day. The purpose of the meeting is to discuss the proposals most likely to be funded, with the goal of putting each proposal in one of several categories. Currently, in OR, SES and MES, these three categories are labeled as follows: primary consideration, secondary consideration and do not consider. Generally, almost all of the proposals placed in the first category will be funded. Occasionally, a proposal in the second category will be funded. Proposals placed in the lowest category are almost never funded, except in extenuating circumstances, e.g., one such case occurred when a proposal was shared with another program, where it was rated highly (placed in their top category), so it was co-funded by one of our programs.

Grantees Conference

CMMI holds a grantees conference approximately every 18 months to showcase the research results from currently funded projects. It used to be that all currently supported CMMI grantees were required to attend, but that has just been changed for the upcoming meeting, where only a (large) subset of PIs will be required to attend; attendance for many who have just started their projects will not be mandatory. More importantly, the conference is actually open to the public, so this is another opportunity to see what type of research is being funded. Attendees can also interact with current PIs to find out the state of the art in various OR fields and application areas. The next meeting takes place July 9-12 in Boston at the Hynes Convention Center. See the Web site at www.cmmigranteeconference.org for details and registration information.

Conclusion

NSF offers many avenues of support for OR/MS research. We welcome input from the community. At the time this writing, the two authors have already announced their intention to finish their terms at two years in late August, so by the time of publication, new program directors may already have been determined. If not, please consider yourself or capable colleagues for the positions. These positions offer a unique opportunity to make a difference in the OR/MS research community. See the NSF Web site at www.nsf.gov for more details. (All of the quotes in this article were taken from the NSF Web site, accessed in February 2012.)

Russell R. Barton (rbarton@nsf.gov) is the Program Director for Service Enterprise Systems and Manufacturing Enterprise Systems, and Michael C. Fu (mfu@nsf.gov) is the Program Director for Operations Research at the National Science Foundation. Any opinions, findings and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.