Runs on light. Easy to park. And you can't see it.

James Tour's custom auto shop is no haven for grease monkeys.

Its walls are white, its equipment glistens, its "mechanics" are gunning for PhDs. And the cars? You need microscopes powerful enough to detect single atoms in order to see them.

Late last year, Dr. Tour and his Rice University lab unveiled the world's first nanocar chassis - an H-shaped molecule with wheels crafted like tiny soccer balls. Now, they've added a motor.

The ultimate goal: to build molecular machines that can mimic nature's ability to transport and assemble tiny building blocks of matter into ever-larger structures.

The project is one small piece of a globe-spanning research effort where "big" is defined as anything larger than 100-billionths of a meter (or about 1/800th the thickness of a human hair). At these scales, materials exhibit unique properties that scientists are trying to harness to build quantum computers; highly sensitive sensors for environmental and medical use; and light, rugged materials for a host of other applications. It's a technological trajectory in which the components become so small that extremely complex devices - say, a Star Trek-like tricorder - may not be too far away.

Biology provides the blueprint for Tour's nanocar work. Key chemicals called enzymes "zip things together. They're nature's little machines," he says.

So the challenge for scientists is to see if they can use small entities such as nanocars to move atoms back and forth predictably and build something, such as highly compact and powerful computer memory chips, Tour says. Eventually, he'd like to have lots of these vehicles "programmed to move back and forth to assemble much larger structures."

That vision moves far beyond products today that claim a nanotech pedigree. Last month, the Woodrow Wilson International Center for Scholars in Washington, D.C., unveiled a new database for tracking consumer products that claim to use some form of nanotechnology - from tennis rackets and hockey sticks using new composite materials to computer chips and sunscreen. Forbes magazine has added a a list of Top 10 nanotech products to its annual set of listings.

Yet these products represent low- hanging fruit - using nanoparticles in sunscreen, for example, to enhance its ability to block sunlight without coating a person's nose with a thick layer of white goo. Tour's effort, by contrast, typifies the challenges researchers face in building the minuscule hardware and vehicles to do the light lifting for bottom-up construction. For all the effort countries worldwide are pouring into nanotech research, many of the tools researchers use are crude, he acknowledges.

His approach to building the chassis is the epitome of mass production. He uses a soup of organic chemicals and processes similar to those used to make pharmaceuticals. The molecular vehicles - the fleet has grown to include trucks and half-trucks - essentially assemble themselves by the millions of billions. The chassis measure a Lilliputian 4 nanometers by 3 nanometers (three-billionths of a meter). It took eight years to produce the first cars, Tour says. It took two years to get images of the vehicles.

Finding an engine, the latest wrinkle, took a relatively quick four months. He and his team have added a paddle-shaped molecule to the crosspiece linking the front and rear axles. When light strikes the molecule, the paddle flips around, much like a riverboat's paddle wheel.

But when the team added the motor, they also had to hunt for a new set of tires. The original carbon buckyballs (geodesic shapes invented by Buckminster Fuller) sapped the energy directly from the motor before the paddle could flip. Such are the unusual effects when driving at the nano scale. The team's results were published recently in two peer-reviewed journals.

So far, the team has successfully tested the motor with the car, in effect, still up on the jacks. The next step is to put it on a flat surface of atoms and see if it scoots under the power of its light-fueled motor.

Selecting the right surface is crucial. And it's a potential two-edged sword. If a metal surface prevents the motor from operating, and the motor can't be "insulated" from the inhibiting effects, the team will have to select a nonmetal surface. But that would require using different microscopes to track the motion - and perhaps lead to less convincing images of what's taking place.

"We have to get the imaging down better," he adds. Still, the showroom is gradually filling. "We're probably going to be coming out with four, or five, or six new models in 2006," Tour says.