FlipSide: A few minutes with... Larry Roberts

Larry Roberts, former chief scientist at the Defense Advanced Research Projects Agency, was in Washington this month to attend events celebrating DARPA’s 50th anniversary. In the late 1960s, DARPA recruited Roberts from the Massachusetts Institute of Technology’s Lincoln Laboratory, where he was experimenting with computer-to-computer networking. Roberts, one of the early developers of the Internet, left DARPA after six years and had a successful career in Silicon Valley as a founder of Internet companies. His latest company, Anagran — “like Anagram with an n” — produces carrier-grade equipment that uses IP flow control technology to manage voice and video traffic over  the Internet. Very quickly, I saw that it was a powerful organization able to do things with very little red tape. If something was substantial and had a big payoff, Congress would support it. We grew the budget by [a factor of]  five from when I started. It was very exciting working with all the researchers in the country. All of them came to me with their ideas, so I learned a tremendous amount. It requires a longer-term view than industry typically has. Some of the venture people are willing to work over a five-year time frame but not a 10-year time frame. In the government, DARPA is about the only agency that is good at innovation. You have to think about what will change things. What I always did was look at the trends: What’s happening to the cost of computing or to the cost of communications? Any sort of trend. It may be the speed of things. Look at the long-term trends. I was doing graphics and virtual reality in the early 1960s at MIT, and then I changed because this is more interesting. All of us involved in the early Internet — Leonard Kleinrock and I, Robert Kahn and Vincent Cerf — have stayed with it because we know it well and have the ideas and all that history. We know what we did wrong. We know what we did right, what we wanted to do and where the Internet might be going. There’s an Emergency Communications and Interoperability Task Force that is trying to get a priority system in place so that an emergency worker in the field can verify himself to the network and get priority access to go through the network. We need to build that capability into the emergency communications system. There was little bandwidth available in the Gulf Coast region after Hurricane Katrina, and you had to use it judiciously. We’ve talked to the [task force] about that. I think it could happen in five years if we funded it and the government moved.