Embedded Files
Tuesday, June 13, 2023
Context: As division director of IIS, I’ve started sending out a short message to the IIS mailing list on the Second Tuesday Every Month (STEM). Here’s the latest installment.
With all the talk about large language models these days, it can be easy to overlook the exciting work that is happening in robotics. NSF is very committed to supporting the robotics community and has a bunch of programs that fund robotics projects. Since robotics is a highly interdisciplinary field, it doesn’t sit perfectly comfortably in any one division, however. Robotics includes systems aspects (CISE/CNS), foundational theoretical work and hardware perspectives (CISE/CCF), mechanical engineering components (ENG/CMMI), electrical engineering engagement (ENG/ECCS), and, of course, perception, learning, and decision-making features (our own CISE/IIS). As a result, it can be a challenge to know exactly where to submit robotics proposals. The webpage https://www.nsf.gov/eng/robotics.jsp (Robotics at NSF) lists almost two dozen programs that are worth considering, along with guidelines about what those programs are looking for.
Arguably the most important robotics program at NSF is Foundational Research in Robotics (FRR). It’s jointly run by the 5 divisions mentioned in the previous paragraph. It’s intended to cover the full breadth of robotics research, but it has some constraints about what kind of work it funds. From the program web page (https://new.nsf.gov/funding/opportunities/foundational-research-robotics-frr), the program only considers work that answers “yes” to all three of these questions:
Is there a robot?
Will a robot gain a new or significantly improved capability?
Is robotics essential to the intellectual merit of the proposal?
If you are doing robotics research that doesn’t answer “yes” to all of these questions, fear not. There are plenty of other NSF programs that could be a match. (See the Robotics at NSF page for guidance.) If you are not sure whether your project fits a given program or not, notice that all announcements of NSF programs include a list of “program contacts” who can help steer your work toward the program most likely to support it.
In addition, there are often webinars and other events where you can ask questions. For example, this Thursday (!), the monthly IIS office hours is dedicated to an “ask anything!” on FRR, hosted by the two program officers that manage the program. The announcement went out last week (June 6th), so please join in if you want to connect with the FRR team.
Last month, I gave an overview of the organizational structure of NSF by likening it to a university. This month, I want to say a little bit about the Merit Review Process, arguably the meat and potatoes of the work being done at the NSF. Here’s a diagram I’ve been referring to a lot lately:
From: https://www.nsf.gov/bfa/dias/policy/merit_review/ .
If you’ve submitted a proposal and served on a panel, you have some experience with steps 1 through 5. You might not know that the next step (#6) is for the program directors involved in the program to discuss highly ranked proposals across panels and come up with a portfolio of projects they would like to fund, informed by the panels’ ratings and the available budget. They write up their recommendation for each proposal in a “review analysis”, which goes to Wendy and me to read (step #7). These review analyses are also viewed by the committee of visitors (CoV), a group that is convened every few years to assess the quality and integrity of NSF program operations. IIS’s CoV is actually taking place this summer, so that’s been very much on our minds lately.
If all goes smoothly and a proposal makes it all the way to be recommended for an award (funding), the last step is for either me or Wendy to “DD concur” the decision, which triggers the process of business review and finalization of funding (steps #8 and #9). In my time at NSF so far, I’ve personally gotten to DD concur roughly $100M in funding. Wendy, who is very adept at the review process, has DD concurred far more. Actually making funding flow to the research community is one of the best parts of the job.
-Michael
Page updated
Report abuse