Frank keeps DARPA on cutting edge

About 25 years ago, Howard Frank wanted to use the communications technology of the time to collaborate on a paper about ARPAnet (Advanced Research Projects Agency Network), the precursor to the Internet. He and his colleague, located in another city, decided to use teletype machines to exchange ideas.

Ironically, Frank and his colleague—both highly regarded information systems scientists—ran into technical snags almost immediately. In the end, Frank said, "we couldn't do it." Exasperated, they gave up on the computers, booked an airplane ticket and met in person to finish the paper. So much for actually using networks to write about network technology.

Recently, as director of the Information Technology Office (ITO) of the Defense Advanced Research Projects Agency, Frank recounted this story to show how far computing has come and to talk about what lies ahead.

Computer developers have tried for 25 years to get things to work, Frank said, adding, "Things broke. They failed." But these efforts finally led to a communications revolution that now allows millions of ordinary folks, not just computer scientists, to collaborate or just chat on-line.

Many of these people use laptop computers and portable modems, without knowing or thinking about the pace of technology that made such communications possible. Frank took a PC Card modem out of his pocket to prove his point. "It's a miracle. It needs to be framed and stuck up on a wall. It's so small you could eat it," he said, flipping it between his fingers with amazement.

Frank speaks with the perspective of someone intimately involved in the evolution of computer and network technology over the last 30 years. He left his position as associate professor of electrical engineering and computer science at the University of California, Berkeley, in the late 1960s to run a technology group within the White House Office of Emergency Preparedness. Later he started his own company and went on to work as a senior information industry executive for 20 years.

Two years ago he joined DARPA as a special assistant, serving as the first director of the DARPA/Defense Information Systems Agency Joint Program Office and then as director of the Computing Systems Technology Office, now merged into ITO.

Frank said the history of IT has been largely devoted to getting "things to work," while the current challenge for computer technology developers, including those at DARPA, is to design information systems that can respond quickly and intelligently to the needs of ordinary users.

"You can see little hints" of these more adaptive, responsive technologies starting to develop, Frank said. Examples include the World Wide Web and natural-language interaction with computers.

Improving human-computer interaction is only part of ITO's overall efforts to develop more agile, adaptive IT. For example, the Defense Department would like to develop ways to draw on the public infrastructures and commercial technologies, such as the growing Web, without exposing U.S. forces to undue risk.

"How do you make use of the commercial infrastructure without really owning it? If you need to control it in order to work, you'll be doomed to failure," Frank said. Ultimately, he added, "you need to be able to react faster than your enemy can react."

Reaction time, more and more, will depend on the ability to construct information infrastructures for particular situations easily. That, in turn, will depend on the ability to make sense of increasingly complicated public and government systems.

"Assembling resources to solve a specialized problem and then dispersing the resources is very, very difficult," Frank said.

Double-Edged Sword

Although DOD has much to gain by tapping into disparate infrastructures and developing systems that are easily and rapidly assembled and disassembled, it also has a lot to lose. "We've created generations of technology, but we've built very little in the way of security and recovery," he said.

"What's worse is we have a very small research community building secure and robust technology. We are the most vulnerable of all nations to attacks of intruders and malicious groups. One of our goals is to create a larger, more robust research community three or four or five years from now," he said.

DARPA's influence in the field of information systems science extends well beyond DOD, Frank said.

"Many of the things we do have vast implications for the nation. It raises the level of IT for everybody. We're adding technology that will give everybody greater capabilities that they wouldn't have had without the research in our program," he said. Just as ARPAnet led to the Internet revolution, DARPA could be sowing the seeds of other IT revolutions right now.