"Houston, we have a problem." Remember those words from Apollo 13? That story still gives me goose bumps - a technically savvy group of engineer-astronauts salvaged a mission gone horribly wrong. What about "Rocket Boys," the story of a West Virginia schoolboy's journey from amateur rocket-launcher to NASA rocket scientist?

This is our cherished science and engineering legend: Talented engineers change the world (usually with the help of caffeine and all-nighters). The story resonates, no matter how often it's retold. Remember "Soul of a New Machine"? The founding of Apple Computer, SAIC, the Internet?

Many of us have come to believe this is the way it always will be: Great engineers will continue to do great things, driving the engine of innovation that fuels the richest economy the world has ever known.

There's just one catch: The fuel supply for that engine is dwindling, at least in the U.S. According to the latest National Science Foundation (NSF) studies, the percentage of scientists and engineering graduates has been declining for decades.

For network specialists, the news is even gloomier. Many of the traditional sources of funding for network research, such as the NSF and Defense Advanced Research Projects Agency (DARPA), are funneling their dollars elsewhere. (I noted in a previous column the critical role both agencies played in funding and operating the early Internet.) According to sources familiar with the NSF, network research is no longer a top priority. DARPA has elected to focus on near-term battle technologies rather than long-range strategic innovation.

You might be asking why any of this is a problem. Why do we need government funding for network research - or any other research - in the first place? Can't industry take up the slack?

Yes, private companies do invest in research and development. Microsoft, IBM, Cisco, AT&T and others operate some of the finest research facilities in the world. But to compete effectively, these companies need to constantly cut out "waste" - expenditures that don't contribute directly to the bottom line. Much great research fits that definition, including the Internet (AT&T actually declined to participate in the early Internet). So relying on private industry to conduct primary research is asking companies to go against their best interests and those of their shareholders - something they're unlikely to do for a sustained period.