27 Nov, 2007
Lightweight, viral systems that grow with use will define how we keep in touch in the years to come
Late last week I spoke with Andy Lippman about the future of communications.
Co-director of MIT's Communications Futures Program and associate director of the MIT Media Lab from 1983 to 2001, Lippman is a deep thinker. And for the past 30 years, Lippman has been thinking deeply about personal communications and computing. As you might expect, Lippman took the conversation to some heady places:
"A lot of people are thinking about telepresence: feeling what is happening remotely...almost like being there."
"When things are connected, all things become opportunities for services.""We will see real-world mashups between people, between people and machines, and machines talking to machines."
But you would be wrong to think that Lippman has his head in the clouds. Holding 11 patents, mainly related to TV and radio signals, Lippman is currently on a year-long sabbatical, working with Nortel's R&D group as a visiting fellow. Moreover, he was deeply involved in creating Nortel's Wireless Mesh Network Solution through MIT.
So what I enjoyed most about our conversation is Lippman's ability to switch between the highly practical, applied science, to the way-out stuff such as telepresence and hyperconnectivity.
"Hyperconnectivity is where the person is fully in the loop," Lippman says. "It will allow you to control stuff on Mars or do surgery from 3,000 miles away."
Funded by commercial companies, the MIT Media Lab came up with the novel idea of creating a consortium. That way, it does not have to apply for funding from each company individually.
"We owe our soul to none of them or all of them collectively," Lippman says, adding that this has allowed the lab to choose research topics that its members feel passionate about.
Of late, Lippman and his students have been working on "viral" communications -- systems that are agile, light on infrastructure, and optimized for invention. He calls such systems viral because, similar to other viral systems, they are built to "catch on person-to-person."
Traditional communications systems are anti-scalable, Lippman says, in that the more people you have tapping them, the more interference you get, and the smaller the bandwidth allocation becomes. Instead, Lippman asks, what if communications systems could grow as the number of users of the system grows?
A fine example of what the good doctor means can be seen by examining the current cellular network. Instead of having a cell phone blast its signal to a distant tower, what if it could jump from computer to computer, or cell phone to cell phone? It would require far less bandwidth and power and could scale organically, because as the number of people using the system grows, the system gains additional connection points.
A simple and elegant idea, isn't it?
And it has a very practical use. The entire one-laptop-per-child idea, also known as the $100 computer movement, will depend on deploying communications in societies where the big, wasteful infrastructure is just not available. But communications will work nicely in "mesh" mode, with each computer passing on data to a nearby computer, using little power and small amounts of bandwidth.
Light and agile, like the doctor says.
I think Lippman's real genius is not in his skills as an engineer but rather in his ability to look at problems from a different perspective: for example, Fluid Voice, the hyperconnectivity technology that he and his students in the Media Lab built.
Fluid Voice is a phone system that acts like a big party line -- push to listen, instead of push to talk.
The default is you can hear everyone in your group who is talking. Instead of a dial, each person is a dot. If you move the dot in or out, the voice becomes louder or lower. Or you can move the dot, and the person, off the screen entirely. You can hear anybody and can decide how much attention you want to pay to each person.
On the practical side, in case of an emergency, firefighters could use Fluid Voice to tune into a police emergency or tune out the ambulance driver, for example.
Lippman's greater point is that the current limitations on radio-enabled communications systems are not due to the physics of the technology but rather to its engineering history. In the old days, radios were expensive, while spectrum was underused and inexpensive. And because receivers were dumb, it became OK to waste spectrum.
Now, the radios are cheap, and spectrum is relatively expensive.
"Mesh is an attempt to break the barrier and history of communications as a limited, fixed, and unscalable system," Lippman says.
"When I went to school, a lot of what you studied was how can you do something given certain constraints," Lippman says. Students were asked to write a program that didn't use more than this much memory, disk space, or processing cycles, for example. Engineering was taught in the context of its limitations.
At some point, engineering courses changed, Lippman says.
"Now students are asked to write a computer program where memory is free, disk space and processing cycles are unlimited," Lippman says, adding that, because of this shift, "we can challenge students to stretch their minds."
Sounds like Lippman is stretching his students' minds in worthwhile directions.Posted by Ephraim Schwartz on November 27, 2007 03:00 AM
