Lost in Translation: Jobs (Hint: We need an ARPA for Commerce)

Tom Katsouleas

The U.S. economy today is more consumer-driven and less investment-driven than at any point in its history. The Obama administration has an opportunity to turn this around by bridging a gap in federal support for research commercialization.

Ever since the divestiture of AT&T in 1984, there has been a steady decline in industrial basic research in the U.S. The role that industrial research at AT&T Bell Labs and others played in innovation and economic stimulus has not been replaced, and as a result, we’re missing a link in the innovation supply chain.  The standard corporate sector response to downturns has become to slash expenses rather than invest in basic R&D for long-term revenue generation.

The U.S. basic research investment strategy has a big hole in it.  Federal funding stops too soon and industrial investment begins too late, creating in between a ‘valley of death’ for new innovations. Valuable research is not getting translated into the products, services and companies that are sources of new jobs.

As a result, many inventions languish in a lab rather than fuel our economy. As an example, it took 50 years for the discovery of nuclear magnetic resonance to come to market as the first commercial MRI machine, a key contributor to human health and the health sector of the economy. We can no longer afford to allow promising latent innovations to languish.

In my own field of particle accelerator research, the U.S. has led the invention of promising new technologies using lasers and plasmas that can miniaturize the size and cost of these sometimes behemoth devices by several factors of ten.  But more work is needed to translate this invention into a product that could, for example, revolutionize cancer therapy with particle beams.  Both Japan and the European Union have development projects to do just that. I believe the U.S. could do it better, but there is no U.S. agency charged with funding the significant translational research needed to advance these devices to the point that private companies could pick them up.

I recently had the opportunity to ask Robert Calderbank, former head of research at AT&T Bell Labs in its heyday and now a Professor of Electrical Engineering at Princeton, about the secret to Bell Labs’ remarkable success. According to Calderbank, the difference between Bell’s approach and university research is simply explained. Bell identified commercial needs up front and then matched them to their capabilities. They came up with a winning technology solution only 3 out of 10 times, but when they did, the result was commercialization nearly 100% of the time.

By contrast, university research funded by the National Science Foundation, for example, seeks proposals from faculty organized along the lines of advancing a discipline such as biological sciences or chemical engineering rather than an application. The success rate is similar to that at Bell Labs and has unquestioned long term value to society, but only a small fraction of successful projects lead to near-term commercialization.

Fortunately, there is something we can do about it. We can fill the void left by Bell Labs by funding the translation of research and education already taking place at universities. Imagine harnessing the real power inherent in 300 research universities and more than 30,000 graduating PhDs in engineering and the sciences each year. 

Universities already do mission-driven basic research, but for the Department of Defense. DOD’s Defense Advanced Research Projects Agency support of basic research proposals from university investigators has led to solutions ranging from detecting improvised explosive devices (IEDs) in Iraq to fighting at the speed of light with laser weapons and predicting the onset of flu epidemics in sailors days before an anticipated deployment.  Energy Secretary Steve Chu and the Department of Energy have taken a page from the defense department’s DARPA program to create ARPA-e, where the ‘e’ is for energy.

What we need now is an ARPA-c where the ‘c’ is for commerce.  Such an agency would identify priorities for mission-driven basic research in areas of critical need for the economy. On a limited scale, the relatively recent NIST Technology Innovation Program is a good start toward this concept.

So how might this work if you are a university physics professor?  Would we all start companies?  Not at all. You might respond to an RFP from ARPA-c just as you might now respond to one from the Air Force Office of Scientific Research for basic research on high power microwaves, except it might now be for basic research on beam propagation and control in tissue (for the cancer application above) or nano-manufacturing control at large scale.  Would it change the culture of university research?  We can hope so and hope not.  Research universities need to provide a rich intellectual environment that accepts and nurtures all types of inquiry, including both traditional research that is not mission-driven and the type here that is. What about national laboratories; isn’t this more aligned with their culture?  National laboratories do have a long and successful tradition of mission-driven research, but generally not the commerce mission here.  Moreover, anecdotal evidence from program managers suggests that universities are a better deal for the taxpayer.  The scale of investment to bring to fruition a paradigm-changing technology tends to be five to one between national labs and universities.

We need innovation to create new jobs now and if we don’t pursue them, other nations beginning to make the right investments will. We don’t have the luxury any more of waiting 50 years for a discovery to become a paradigm-shifting new economy.  The world won’t wait with us.

Tom Katsouleas
Dean, Pratt School of Engineering
Duke University

This contribution has not been peer refereed. It represents solely the view(s) of the author(s) and not necessarily the views of APS.