What's next for computers at Carnegie Mellon

Related Images

Enlarge

Enlarge

Enlarge

Enlarge

Related Tags

Oakland 

If the next 50 years in computers are anything like the last 50, hold onto your virtual hats--or in this case, heads. Anything is possible.

It was only five decades ago that the computer age began, at what was then Carnegie Tech, when Professors Herbert Simon and Allen Newell invented “a thinking machine” the size of a garage. In hindsight, the enormity of that first computer is symbolic of how large-scale the invention of artificial intelligence would become in transforming our world.

This month Carnegie Mellon University celebrates with CS50, a series of events scheduled April 19th through 22nd, that highlights the stunning progress of computer science in the last half century.  

In the beginning, “the original computers took up huge room with kilowatts of energy,” says Dr. Randy Bryant, dean of the School of Computer Science at Carnegie Mellon. “Now we’ve gotten to the point of laptops and iPods and cell phones, small personal devices. If that projection continues, computers are going to be microscopic in size and you’ll have a thousand operating in the size of a breadbox.”

Once computers become that small and powerfully active, we have to rethink them altogether, he asserts.

“A lot of the progress to date has been driven by hardware getting smaller, faster, and cheaper,” Bryant says. “The main position we’ve taken is to think broadly about what computer science can mean--beyond building and programming, to all the ways computers can be out there doing useful things such as human interaction.”

One likely and appealing scenario? “A whole city with computer sensors monitoring traffic, preventing accidents and sending drivers on different routes,” says Bryant. Give it twenty years, he predicts.

Heads Up - and Out

The most out-of-the-box thinking, says Bryant, is nanotechnology dubbed Claytronics that could recreate a 3-D structure in a remote site--say, your home office--from teleoperations at the source--let's say corporate headquarters. “It has interesting teleconferencing capability; you could see a person’s head on a table (instead of a screen) in 3D,” Bryant offers although even he admits, “It would take getting used to.”

If you find that hard to envision, try this: that same glowing 3D whole-body replica of a person, with light-emitting diodes and photo cells on its electromagnet surface, that allows it to move, touch and see. In the ultimate conference call, these nanotech robots could be built for each person in an organization for meetings that are always, and perhaps eerily, face-to-face.

Or consider a space-age house call from your physician. The lifesize replica, controlled remotely by the doctor, would be right there with you, talking, probing and examining you. For this to work, a replica of you would be at the doctor's office so the actual doctor would be examining, in real time,the nanorobot recreation of you. How wild is that--and how is it (remotely) possible? Through catoms, or Claytronic atoms, billions of tiny computers in one robotic recreation. When the exam is over, you and the physician would "disassemble," leaving behind a pile of Claytronics. 

This is, literally, out of the box thinking that takes modular robots to the extreme. Imagine video figures appearing--in action mode—right there in your great room instead of in a confining television. Or the ability to program shapes and designs with these catoms, and morph a chair, for instance, into a sofa. "The end result is very cool," says Carnegie Mellon assistant professor Seth Goldstein, "but the essence is trying to manage complexity management--a billion computers working together to do one task. And that's the science behind it."

While these inventions are "long term visions," six core researchers and 20 members of the Carnegie Mellon faculty, along with Intel, are already at work on the technology.

Hard to grasp it all? Think back a century or two to how someone might have tried to fathom the concept of watching television. Let alone viewing it on a portable wireless device called an iPod.

"It takes awhile to metabolize," agrees Goldstein, who will be demonstrating catoms at the cs50. "But it is fantastic."
 
You Can Call Me Pearl

Much closer to reality, an easier to grasp technology comes in the form of the pert, red-lipped Pearl the Nursebot, a joint project between the University of Pittsburgh and Carnegie Mellon which is currently being field tested in Pittsburgh. Designed to assist your elderly relative (or, sorry, maybe even you by that time) in living independently, Pearl will remind patients to take medication, collect data and act as a tele-presence for doctors to interact directly with remote patients. 

“Think about the value of elderly people living independently. The economic impact is tremendous since the cost of living is so high,” Bryant asserts. What’s more, that technology doesn’t have to be exotic, he adds. Videoconferencing is a sound option to communicate with physicians or patients, and help the infirmed or elderly or shut-in when they most need it.

While the technological progress dazzles, at the same time it results in a troubling paradox: as technology advances, developing countries only fall further behind. Although many groups are helping these countries, few are figuring out how to help through technology. Carnegie Mellon is at the forefront here, too, with the TechBridgeWorld initiative. One of their goals? To narrow the digital divide in a number of ways: robots for use in removing land mines in war-torn countries, telemedicine for medical support to communities that lack healthcare, or sensing and computer technology for safety and quality in industry.

Act Global, Act Local

On the local front, one thing Bryant would like to see in the future is even more collaboration between Carnegie Mellon and the University of Pittsburgh in “building up our techno-economy in Pittsburgh. It’s not nearly as successful as heavyweight competitors Stanford or MIT, schools that have achieved greater success in collaborative community projects. Yet, “it’s important to Carnegie Mellon as well as Pittsburgh,"he says.

But progress is being made. Getting Google to locate in Pittsburgh was a big step. “Think of it as the anchor tenant in a very collaborative and interactive high-tech environment,” Bryant suggests, adding that the School of Computer Science is talking to “every company we can.”

These high-tech jewels in the crown have great potential for local collaboration, such as the recent one between UPMC and the Intel lab at Carnegie Mellon. Although Intel was originally set up as a connection solely for Carnegie Mellon, the two organizations recently signed an agreement with UPMC’s medical information systems and processing to manage the "tons of  X-rays" it has stored away. Although it was a perfect fit, “it wouldn’t have happened directly between UPMC and Carnegie Mellon,” says Bryant. “It took the wiring of Intel to form that connection.”

Sounding the Meds and Eds theme, Bryant says, “Computers and medicine have the potential to make Pittsburgh stand out in a unique way for the country. We could be better than Boston, particularly in the use of medical robotics.”

That's one thing, perhaps, that's not hard so to imagine.

Tracy Certo is editor of Pop City.

Photos:

Catoms and Floppy Disks

Professors Simon and Newell

Randy Bryant, Ph.D

TechBridge Logo

Early computer