Jason Campbell walks down the hall, heading toward the conference room. With his wire-rimmed glasses, ponytail, and backpack, those who glance too quickly might mistake him for one of his students. He’s curious about the upcoming staff meeting. Most of these are routine, but this one sounds, well, a little crazy. He opens the door and stops, surprised. The room, for a voluntary meeting, is unusually full. Apparently, he's not the only one intrigued by what the new director and his colleague have to say. All 20 researchers from the Intel Research Pittsburgh lab at Carnegie Mellon University, along with many of the grad students, sit waiting. Campbell, a senior researcher and adjunct university staff member, finds one of the few remaining seats. Sitting in a nondescript conference room above a Starbucks near campus—who would think this was where he’d learn about a technology that could change the world?
A few years before this staff meeting, Seth Goldstein and Todd Mowry, both Carnegie Mellon computer science professors, happen to attend a Computing Research Association conference. Mowry is interested in new technologies to help people communicate. By chance, he reads Goldstein's conference paper on reconfiguring the very molecules of a computer. Mowry's imagination fires. During a break, he grabs his colleague for a chat.
They talk about communication. Cell phones. Video conference. Skype. They talk about computer molecules and the incredible concept of shape-shifting matter. They pause. Maybe molecules will come later. Maybe they should start with particles more like ping-pong balls. They talk about phone calls going beyond 3D if that's possible. By the end of the conversation, they’re talking about communication like no one has ever experienced.
The ramifications are infinite. Imagine a Pittsburgh-based UPMC surgeon orchestrating a heart transplant in Qatar by using a moving, lifelike cloned image of the patient. Or a four-star general strategizing overseas military maneuvers with tiny, morphing troops mobilized on his Pentagon office desk. Or even a Yankees fan watching his miniaturized team turn a double play against the Red Sox right there on his kitchen table in the Bronx beside a cold piece of pizza. As Goldstein says, "It's like being there without being there."
While Goldstein and Mowry brainstorm at the conference, Campbell is in California, after graduating from Stanford with a degree in electrical engineering. He's not interested in run-of-the-mill corporate jobs. He's chasing start-ups where he can be "doing something nobody's done before." He's intent on discovering "what's it gonna take for computing to transform our experience as people," a bold statement, even from a budding technology entrepreneur.
Meanwhile, Intel, a Silicon Valley microprocessor giant, is looking toward Pittsburgh. The company that designs and builds the brains of computers knows that its chips must be able to handle the next big software application, whatever that might be. One way to have such a crystal ball is to work with some of the best research faculty in the world. For that reason, Intel establishes a collaborative research facility near the Carnegie Mellon campus, not far from the offices of Goldstein and Mowry. The unusual thing about this corporate lab is that it's not a slave to patents and the bottom line. The real focus stays on knowledge—creating and sharing scientific breakthroughs.
As the new Intel lab takes root, Campbell, by chance, relocates to Pittsburgh, where his future wife is studying for her PhD. His aspirations haven’t changed. He wants to work on robotics, and he wants a place willing to gamble on risky ventures with transformational potential. He thinks the Intel lab looks like the perfect fit. Campbell signs on and soon finds himself working for a brand new director—Mowry.
The lab is collaboration. Graduate and undergraduate students work fluidly with both Intel researchers and school faculty, and Intel staff co-teach university classes. Now at the helm, Mowry is anxious to generate interest in his futuristic project among the best and brightest in the offices around him. An Intel lab director doesn’t just march in and dictate. This lab is about ideas. He wants to attract scientists who are passionate about the research. It's the only way to move a groundbreaking project like this forward. Mowry and Goldstein call the staff meeting to introduce what they first discussed at that conference—combining the images of claymation, a form of stop-motion animation using clay, with the power of electronics.
Campbell and the others in the packed conference room look up expectantly. Goldstein and Mowry launch in. They describe their vision of shape-shifting matter and telepresence that they dub Claytronics. The mood remains interested but skeptical. They’ve all heard fantastic, impossible ideas. Campbell glances around and silently thinks he's "probably the most skeptical in the room." With what they're describing, Campbell believes "there’s a 50-year time frame."
Fueling Campbell's doubts, the professors explain that using a single material, as we think of modeling clay, isn’t the ideal way to achieve their goal, despite the catchy project name. More promising would be a method in which a number of discrete parts they call catoms—short for claytronic atoms—would rearrange themselves to form different shapes.
Campbell describes it later as similar to concrete, where particles are held together by glue, except in this case the glue would be controllable and the grains would move themselves around. The grains, in fact, would be miniature robots, the size of pixels on your TV, each with the capability to compute, move, and communicate. They would act like cells in the body, not separate, but dependent on each other for movement. "If you set one in the center of a table, without a neighbor to work with, it can't go anywhere," Campbell explains. "If you give it a couple of friends, then they’ll be able to work their way across the table together."
Mowry shows the assembled group the catoms they’ve created to-date—two cylindrical objects the size of typical restaurant saltshakers, each with three magnets attached around the outside. At this point, if pumped with lots of power, they’ll jump slightly.
Hands shoot up with questions, none more often than Campbell's.
How can you supply enough power to make this work?
Won’t there be issues with getting rid of the heat produced?
How would you keep the catoms from sticking together?
How can you feasibly envision a few saltshakers becoming millions of miniscule, shape-shifting robots?
There are no simple answers, only more discussion. These are, after all, just the sort of challenges Goldstein and Mowry hope bright researchers like Campbell can help them solve. And it appears they have Campbell thinking.
The meeting ends. Campbell remains skeptical, but he can’t get Claytronics out of his mind. He is, after all, hooked by radical, risky ideas—something that just might change the world as we know it. Something that hasn’t been done yet, maybe not even considered, "because if it's done already there's no reason to join. You're a researcher, you want to make something happen!" he says.
Over the next few months of thinking, talking, and considering possible answers to his questions, Campbell sways from skeptic to believer. He goes so far as to make a striking comparison—Claytronics could prove as revolutionary as the transistor, which is the key component in almost all of today's electronics and is referred to as one of the greatest technological breakthroughs in history. He becomes Intel's principal investigator on the project while Goldstein maintains the lead for the university.
Five years have passed since that first meeting in Intel's old conference room.
The lab has moved on campus, making the physical side of collaboration even easier. They were the first to occupy a brand-new Carnegie Mellon building dedicated to joint industry/university research. According to Campbell, it was aptly named the Collaborative Innovation Center for its first tenant. The office configuration is nearly wall-free, to purposely encourage communication between staff and the university. Campbell, pointing outward from a glass-enclosed conference room, jokes, "We can just about see Seth Goldstein's window in Wean Hall. If I'm in his office and I stand next to the window, I can get my Intel Wi-Fi network."
With the relocation complete, joint research continues. As expected, Campbell finds creating the hardware and software necessary for radical technology more than just a little difficult. Goldstein describes the overall problem as one of "scale." To make miniature robots less than one millimeter in diameter, they have to create hardware scaled way, way down. On the other hand, to get a million of those tiny computer-robots working together simultaneously, they have to create software ramped up beyond anyone's imagination. To make this happen, dozens have come in contact with Claytronics. Campbell and his colleague Babu Pillai have been the primary researchers from the Intel side, while scores of faculty and students from computer science and electrical and computer engineering have taken part from Carnegie Mellon.
There has been plenty of excitement. Working toward "ensemble" programming, the team has developed two new powerful programming languages that strongly hint at success. Magnets didn’t work, but electrostatic energy has, which Goldstein compares to the child’s game of rubbing balloons so they stick to a wall. The original saltshakers have shrunk, grown, then shrunk even further. They're now miniscule tubes, just a millimeter wide by 10 millimeters long, smaller than a chunk of a toothpick. The early catoms jumped, and now they roll, but without integral circuitry or intelligence. Now the researchers are in the process of building the circuitry right into the catom itself, and they are confident they’ll soon be demonstrating the mechanisms working—showing that catoms can be controlled, move against gravity, share power, and communicate.
Once that's accomplished, Campbell believes we'll see a lab demo of something like a cell phone that, with the push of a button, can grow into a laptop computer or shrink to an MP3 player.
Goldstein likes the shape-shifting application, but he offers another possibility. Picture a surgeon ready to remove a precariously located tumor. Instead of a scalpel, he picks up a syringe. He fills it with catoms and injects them as near as he can to the growing tumor. The catoms flow in, one by one, sending a stream of images back. The surgeon types in instructions, and the catoms surround, grow, and destroy the diseased tissue, then file out of the body. Claytronics could turn complex surgery into something akin to getting a chicken pox vaccine. With fewer catoms needed to fill a syringe than a growing-shrinking monitor screen, Goldstein thinks this medical breakthrough may happen first.
There is another byproduct of developing communicating catoms. As society has become more dependent on interconnected computer systems like the power grid, air traffic control system, water distribution system, and the Internet itself, we have created what Goldstein calls "fragile, brittle" systems that we don’t fully understand. A defective link in the chain can send a whole system crashing, as in the massive northeast U.S. blackout of 2003. Goldstein believes that the technology developed for Claytronics can help us begin to understand programming units as true ensembles.
All of the Claytronic team members agree that it’s impossible to predict what others might invent with this technology. They also agree that the technology isn't that far away, certainly not the "50-year time frame" Campbell estimated when he first questioned Claytronics technology. He now believes that within five years there will be moving, programmable, millimeter-sized catoms that will be ready for everyday applications. "I believe that so strongly that I feel if we [Carnegie Mellon and Intel] don't do it, someone else will," he says.
Goldstein goes further. While it's easy to create headlines describing Claytronics as a really cool cell phone or medical marvel, he says it’s so much more. "Understanding the complexity of programming millions of cooperating computers is, I think, the fundamental computer science challenge of the next 20 years. If we succeed," he quietly says, without a hint of irony, "it will really change the way humanity operates."
Melissa Silmore (TPR'85) is a Pittsburgh-based freelance writer. She is a regular contributor to this magazine.
