In Defense of the NSF


In 2009, Mont Hubbard, Ronald A. Hess, Dale L. Peterson, and I were awarded $300K for a two year grant from the National Science Foundation (NSF) to study how people are able to balance on bicycles. We completed the work in September 2012 after an additional one year no-cost extension. This resulted in numerous research products that have been well cited since the completion of the work.

In 2011, US Senator Tom Coburn (R-OK) released a report entitled "The National Science Foundation: Under the Microscope" with the intention to expose wasteful spending of US tax dollars by the NSF. The grant we were awarded was listed as one of the examples of "silly research". This received a fair amount of media attention and we were interviewed for a number of publications to comment on this report, e.g. in IEEE Spectrum.

Then again in 2014, US Assemblyman Lamar Smith (R-TX), the chair of the House's Committee on Science, Space, and Technology, launched an investigation into NSF's wasteful spending, also bringing our grant and work back into the spotlight.

My colleagues and I have responded to various media inquiries but never developed a statement in our own words that reflects our opinion on this matter. Now 4.5 years later, this past week a journalist contacted us with this question:

Hi Professor Hubbard,

I'm a reporter with <redacted> writing on the implications of the Scientific Research in the National Interest Act (HR 3293). I'm highlighting interesting research that lawmakers decried as a misuse of federal funds and came upon your NSF-funded research about how people ride bikes. I was wondering if you knew of any instances of your findings from that study being used in a meaningful way, perhaps by cities to decongest roads and increase bike ridership or something along those lines. I'd love to include it in my piece.

Thanks for your time!

Best,

<redacted>

This request prompted us to formulate some responses. In particular, I finally found a moment to put some thoughts to paper. The letter below is my and Mont's response to this journalist that we believe summarizes our opinion on the continued attacks by some Congress members on the NSF:

<redacted>,

Thank you for your inquiry. Yes, our NSF funded research has been targeted in the past by some members of Congress that consider it to have been a waste of money. You asked if we "knew of any instances of [our] findings from that study being used in a meaningful way, perhaps by cities to decongest roads and increase bike ridership". This question suggests that you may have the same misunderstanding about our research as some members of Congress do.

The primary purpose of our research was to develop experimentally validated mathematical models that accurately predict the interaction between, and motion of, a human and the vehicle they are controlling. We chose the bicycle because it is one of the more dynamically complex vehicles humans learn to control and because it is a familiar, inexpensive, and tractable experimental platform (e.g. it is orders of magnitude cheaper than buying a F-22 Raptor and performing experiments with it, yet you get similarly meaningful results).

In the four and a half years since our NSF grant ended, our research has influenced the robotics, vehicle dynamics, manual control, vehicle safety, sports engineering, biomechanics, and system identification research communities, which can be seen by the more than one hundred of citations to our research products. Additionally, the results of our NSF funded work have played roles in the progress of a number of companies, some of which have employed the 10+ graduate and undergraduate students that were involved the NSF funded research. We also speculate that a long-term, broader impact of our work may specifically affect future bicycle, scooter, and motorcycle designs that could contribute to our country's transportation needs.

But, as with many of the congressionally targeted studies, there seems to be a fundamental misunderstanding of the purpose of the questioned research. NSF requires applicants to write a broader impact statement and this ends up as fully a third of the publicly visible grant abstract, whereas it is often only a very small component of the work and deliverables specified in the full proposal. This abstract is what a congressional aide may typically read. We do believe that our research will eventually play a role in the broader impacts we wrote, but it is much too soon to know for certain and further work must be done to ensure the impact. Also, as with any research, we may be wrong and our broader impacts may not come to fruition. It would not be called research if that risk were not present.

One story that needs to be written is about the NSF's inability to properly communicate science to the public and Congress and the corresponding misunderstanding of science that the public and Congress seem to have. The scientific process is a remarkable one that thrives on vehemently eliminating untruth. Because there are only a few scientists in Congress, little to no experience with this process is present. This lack of experience contributes to the lack of trust in the funding, results, and reliability of science. NSF could obviously be doing a better job convincing Congress of its merit.

Another story could include the fact that the broader impacts section of NSF grants is often a meaningless statement to fulfill a checkbox that is rarely actually delivered. For example, many robotics grants cite that their robots will help in search and rescue missions in natural disasters, just to pander to the popular sentiment that this is a good use of research funds. Yet few robots are developed for that specific purpose. You could just count the number of papers and grants that claim this to find out if the robot ended up in any actual applied search and rescue missions.

Obtaining a NSF grant is a highly competitive process. Only about 20% of proposed grants are awarded. Once the NSF establishes its research directions, the competitive review process is one of the most thorough to be found, in which national experts are gathered to critically screen the submitted proposals over several days. We agree that there may be some small percentage of past NSF proposals that this careful process overlooked that might have eventually benefited society more than the ones chosen. But we feel this percentage is so small that it does not provide any evidence that NSF's system is poor or broken. We have high confidence that virtually all funded NSF research grants are worthy of being funded.

Funding scientific research is not like going to Walmart to buy a "whatzit". Our present unfunded research has continued along the lines originally begun in our NSF grant. Specifically, we are now researching the process of optimal design of bicycles (which actually is something of a whatzit). This present research was enabled by the prior NSF grant and is an entirely meaningful contribution resulting from the NSF's support.

We sincerely hope that your article will highlight the disconnect between the House Committee on Science, Space, and Technology and the scientists and engineers that NSF funds. Just producing a tit for tat exchange on whether their targeted funded grants actually produced something is missing the most important point.

Sincerely,

Jason K. Moore, PhD, Faculty*
Mont Hubbard, PhD, Professor Emeritus*

*Mechanical and Aerospace Engineering, University of California, Davis