|In Defense of DARPA; Lamenting Bell Labs
|By Barry Manz
A federal agency like DARPA is a sitting duck for politicians and assorted other critics. It has come up with some truly bizarre programs over years that ultimately either delivered no tangible results, were canceled before they could cause any damage, or attempted to answer questions that nobody was asking or needed answers to. You can find a few of the most notorious in Table 1.
However, this is basic research, which is always risky: Some projects will produce results but most will not, a scientific equivalent of day-trading or casino gambling: You only have to be right once. In that context, DARPA is winning big:
• The Internet: DARPA’s predecessor ARPA created it.
• GPS: DARPA’s five-satellite Transit family provided the first satellite-based geolocation capability with resolution of 200 m.
• Stealth: It began in part with a DARPA program to create low-observability aircraft, which resulted in the Have Blue project and the first production stealthy fighter aircraft (Lockheed’s Skunkworks-spawned F-117).
• GaAs: DARPA’s MIMIC program turned a fledgling technology into a marketable semiconductor material that later created an entire industry. The cost was a paltry $700 million.
• GaN: Without DARPA’s development money and DoD’s push to get GaN into the field, this semiconductor material would not be at the stage it is today.
• Speech translation: Currently being used with some success in Iraq and Afghanistan, this DARPA-championed technology will ultimately prove essential, compensating for the lack of human translators.
The list of DARPA programs that have “worked” is a long one. How much economic benefit can be ascribed to its efforts is difficult to determine but undoubtedly substantial. Its work in 2013 will consume less than $3 billion of taxpayer money -- lunch money in DoD’s $700 billion budget and microscopic in the $3.8 trillion federal budget.
So even though DARPA often misses the mark, any prudent investor would consider it a great return on investment, and DARPA does this work in partnership with private industry, it speeds up research. But its most important contribution is filling the hole where private industry basic research used to reside.
Bell Labs: A Tradition Sorely Missed
Bell Laboratories was perhaps the world’s most prolific innovator, with a list of inventions that is breathtaking to behold. The short list is shown in Table 2. Yes, Bell Labs got government money, but the organization spent billions of private money (i.e., its own) developing technology that has in considerable measure changed our lives. Like DARPA, its goal was to break new ground through basic research that more often than not didn’t produce results. I once knew a microwave physicist who spent most of his career at Bell Labs in Murray Hill, NJ, and produced some interesting advances, but he was the first to admit that most of his work came to naught.
It’s all gone now, thanks to market pressures, in particular the 2006 Lucent Technologies merger with Alcatel, but its days were numbered even before this. In 2007 what was then called Lucent Bell Laboratories and the former Alcatel Research and Innovation were merged into one organization once again called Bell Laboratories. By 2008 there were reportedly only four scientists left in physics research and later that year Alcatel-Lucent abandoned basic science, material physics, and semiconductor research entirely to focus on technologies on which it could presumably make money more quickly like networking, high-speed electronics, wireless networks, nanotechnology, and software.
Few publically-traded companies today either can afford or have the interest in pursuing basic research that may or may not pay off. Of necessity they focus on developing technologies that are likely to deliver results in the short-term and keep shareholders happy. The trend toward taking companies private has as much to do with being able to develop technologies and products without shareholder scrutiny as it does with dealing with the expensive annoyance of Sarbanes-Oxley. Few electronics technology companies short of powerhouses like Apple and Google have pockets deep enough to explore more than one basic technology at a time.
Fortunately, organizations such as the Xerox Palo Alto Research Center (PARC) still exist, but they are few. Academia provides enormous resources, albeit of a different type, to industry and government. Beyond these two categories, we have only the government, in the form of DARPA, the National Institutes of Health, and other non-defense agencies to tinker with basic research in partnership with industry. Only the most ideologically rigid would try to make a case for eliminating this type of cooperation because quite simply it produces results that can create new market and even new industries -- which translates into new companies, greater productivity, and higher employment.
I’m not suggesting that blindly throwing huge sums of cash at companies like Solyndra is sensible. It was instead stupid, politically motivated, or both. But compare that with DARPA’s penchant for throwing down the gauntlet to industry to leapfrog over currently technology to reap rewards orders of magnitude greater – even if many fall flat – and the difference is clear, at least to me.
Barry Manz is a contributing editor to Microwave Product Digesting can be reached at firstname.lastname@example.org.