3D You

A team of researchers may bring the ultimate sci-fi fantasy to life.

by / January 29, 2007
At the scene of an automobile accident, several vehicles are engulfed in a raging inferno. Firefighters and other rescue personnel arrive on scene and receive a report that multiple victims are trapped in the flaming wreckage. Hundreds of gallons of water are quickly expelled to extinguish the blaze. In all likelihood, the vehicles' occupants are dead. Nevertheless, firefighters examine each smoldering heap and find one person with faint vital signs.

Rescuers begin prying the car doors, breaking them free from their chassis. Then, in what seems like a magician's sleight of hand, the firefighters' tools evaporate only to re-form moments later into a stretcher. The victim is loaded onto a spectral, makeshift bed and rushed from the scene to a waiting ambulance.

The captain barks an order -- the drill is over. The firefighters breathe a sigh of relief, and drop the stretcher and its occupant to the ground, where they appear to shatter into a billion tiny pieces.

Welcome to the world of 3-D holograms. But these holograms, officially known as "dynamic physical renderings," are not merely figments of light and color -- they have mass, weight and texture. They move in real time and interact as if they were actual physical objects -- even actual people.

If it seems like science fiction, it is. But for a team of researchers at Intel and Carnegie Mellon University (CMU), science fiction becoming science fact may be far closer to happening than anyone imagined.

Taking Shape
Just over three years ago, a couple of computer science/engineer types found themselves at a brainstorming session and wound up with the next big idea.

Todd Mowry and Seth Goldstein, both associate professors of computer science at CMU, hit on an idea that could fundamentally change the world.

Mowry imagined a technology that would let people project what he calls a "telepresence" -- a remote, three-dimensional representation of a human being. The representation would not be merely an image, but a physical duplication or model. The technology would, for example, replace telephone and Web conferencing by creating lifelike replications of the conference participants, all in the same room.

"Seth and I came up with the idea for the project," Mowry recalled. "We were at a workshop sponsored by the National Science Foundation (NSF) and Computing Research Association, where we were supposed to be brainstorming about big, grand challenge-types of ideas.

"Seth had a proposal for using possibly nanotechnology, but not necessarily that, to build little objects like robots that could form into shapes. We sort of realized the best way to build what I had in mind [with telepresence] was through Seth's idea of having things form into physical shapes -- to have something that is physically there, rather than the illusion."

Mowry and Goldstein were convinced they were on to something, and believed the technology existed to build something they called a "claytronics atom" or "catom."

When one very small catom is combined with billions of others -- along with some powerful software -- this amalgam could be programmed to take the physical shape of whatever a user wanted.

CMU provided initial funding to investigate the idea's feasibility. The notion soon attracted others who saw the potential.

Jason Campbell, a senior researcher at Intel Research Pittsburgh, joined Mowry and Goldstein on the project. Additional funding now comes from Intel, the NSF and the Defense Advanced Research Projects Agency. As research progressed from idea to prototype, Mowry took a leave of absence from CMU to serve as director of research at Intel Research Pittsburgh.

After three years of work, the team has produced several simulations and larger-scale prototypes of catoms. A catom is an individual unit, much like an organic cell, that when combined with others, forms an object. Also like a cell, it needs power, programming and a cohesive force. The human cells use energy, are programmed by the brain and bound together by various forces, such as electromagnetism or chemical bonds.

"In terms of the hardware side, we're starting to build prototypes at a couple of different scales -- some very large prototypes that replicate all the computational and other components we want. Those are about two inches across," Mowry said. "And some very small prototypes in the submillimeter realm that are just starting to replicate geometry. We expect, in the coming year, to start to put transistors on those as well, which would give us the computation we need."

If all goes as planned, the team says, one of the earliest applications might be only a few years away. They call it the 3-D fax machine.

Using catoms Goldstein predicts would be a millimeter in diameter -- still very large but functional -- users could capture and reproduce any arbitrary object using a new breed of fax machine, much like a sci-fi teleportation device.

Unlike teleportation, however, the machine would merely duplicate an object as opposed to actually sending it to another locale. Ideally if the technology is perfected, the catoms would be so small they could emulate any texture. But Goldstein is optimistic about the 3-D fax machine, despite its bulkier catoms.

"We have looked at a lot of other applications, and probably the most convincing one to me in terms of both how close it is and also how much work we've done on it, is this idea of a 3-D fax machine," Goldstein said. "The reason that seems so close to me is that you don't actually need to have the thing moving dynamically in real time. The hardware mechanisms you need to build that are fairly simple."

By "dynamic movement," Goldstein means that this 3-D fax machine would not have to worry about duplicating objects that move independently -- instead the replications would be stationary, inanimate things.

The catoms -- the hardware -- are one of the two primary challenges of making dynamic physical rendering a reality. The team believes that building the catoms is, while certainly a tremendous engineering challenge, "eminently doable." In fact, Campbell recalled, the first great leap forward from mere idea toward reality was when they got a prototype to move.

"It took off dramatically when the very first prototype wiggled -- which was still very far away from a working catom -- and we were able to show that to people. We then got the interest of a lot of faculty. Then suddenly the project grew significantly in terms of the number of CMU people involved. Another big leap was when the project became not just a CMU project, but an Intel project as well. In the last year, we've made a fairly huge amount of progress."

The team has made some advances on the hardware side, experimenting with various adhesive forces such as electric fields and electromagnetism. Currently the researchers say that electric fields hold the most promise on the scale of microscopically rendered objects. The challenge is conducting experiments on a microscopic level to confirm their suspicion. Such experiments are difficult and costly.

Regardless, the team is quite confident that, with the "proper investment," the catoms can be built. If anything derails the project, it won't be building the catoms, it will be designing the software. The software will be the brains of an object, telling each catom how to move, what color light to emit, what arrangement will result in the proper texture, among other variables.

If the object were a human replica, the number of catoms the software would command would be in the hundreds of billions.

"On the software front, we've made great strides in the last year and a half or so -- identifying ways to form shapes and route power and start to control some of these devices -- but we have a long way left to go," Mowry said. "The software is a key challenge. In fact, the software is more difficult than the hardware."

Trek Tech
For fans of Star Trek: The Next Generation, catoms, 3-D fax machines and dynamic physical rendering sound a lot like the grand fantasy playground that is the starship Enterprise.

On the TV series, crewmembers frequented what they called the Holodeck, a room where lifelike holograms engaged users in whatever scenario they imagined. Star Trek fans are known to pine for the gadgets and gizmos imagined by the show's writers -- none more so than the Holodeck.

"It's not far off," Mowry said. "In the case of the Holodeck, you'd imagine constructing an entire interactive environment. At least initially we think of having similar environments in multiple places, and then constructing a replica of the person with whom you are having a conversation. Ultimately it's along the right lines."

As the researchers continue working, they must figure out pesky issues, like how to power the things, how to get the hardware down to the desired size, and a million other details. Fortunately they have ideas about how to solve most of the problems. A solution currently in play is to power a vat in which the catoms are stored, with a direct connection between the vat and the finished object.

Since the beginning, they've referred to the rendering of catoms as claytronics -- essentially the molding of electronic "clay" into the most realistic artificial replication possible. Although the final product is a long way off, the name is especially apt considering Mowry says all it will take to reach the next level is molding a simple geometric shape.

"Once we have four real units, and one of them can crawl on top of the other one and form a pyramid, I think you'll see a tremendous number of groups want to get involved," Mowry predicted. "I think our expectation is there isn't one group that is going to solve this problem, but rather we are starting an area in which many people will get onboard once they realize it's as doable as we think it is."
Chad Vander Veen

Chad Vander Veen is the former editor of FutureStructure.


Platforms & Programs