A recent Friday at the Second Decade Society room at Johns Hopkins University witnessed the competition of nine robots to see which could produce the best painting on a large square of paper taped to the ground. Student engineers and artists mingled over food and drinks, following the judges from one display to the next and crowding around to see the 'bots scurry and plod across the floor. The crowd was forgiving of technical difficulties--there were a few--if a bit cautious. "It's just spraying paint wherever it goes," the creators of one robot told the audience. Everyone took a big step back.

Early favorite Can Dinsky was a spray paint-loaded hexapod named for Wassily Kandinsky, a Russian painter and early proponent of abstract art, who wrote in 1911 that "every work of art is the child of its time." His namesake had stalklike antennae sensors and a sonar device on its back that activated a paint syringe in its belly. Can Dinsky's six Plexiglas legs, which elevated it above its wheeled competitors in complexity, ultimately proved the robot's undoing. One rubber foot or another became stuck in the drying spray paint as it traversed the canvas. Thanks to the noxious cloud of paint fumes that filled the room while the artist worked, it took a prize for "Most Dangerous."

The people's choice award, which audience members decided by affixing little blue dots to sheets of paper in front of the robot displays, went to a wheeled box called, for no particular reason, M&M. Its style was more strictly geometric than Can Dinsky's. It relied on a pair of geared circles modeled on the old Spirograph toy design, holding two colored Sharpie markers. A circuit strapped to its back generated random numbers to determine how many spins of the Spiro were made in a certain spot and how far the machine would roll forward before making more spins. It was constructed over "a couple of weeks," says Alican Demir, a Hopkins undergraduate from Turkey who was part of the two-person team that designed and built M&M, "the last week being very intense." The pair had been up most of the night before working out the kinks.

The contest was judged by a group of artists, who doled awards to each team. "When we see things," offers judge Joe Reinsel, a digital artist who teaches at the University of Maryland, "it's our own perception that actually makes it the art, not the piece itself."

"It's the process of making it," counters judge Joan Freedman, the director of Hopkins' Digital Media Center. "There's some kind of intellectual process going on in the making of the art, and that's what I think these robots, with the level of engineering [their creators] are willing to do--I don't think they can make robots smart enough to make art."

"There is a person involved in making the art," Reinsel says. "So in a way it's the person making the art."

The person ultimately responsible for this art is roboticist Allison Okamura. The Hopkins professor of mechanical engineering assigned the project to undergraduate engineering students in her mechatronics class.

This is the second competition of robot artists overseen by Okamura. Last year, when she took her maternity leave, the artbots didn't happen. "Allison is too shy to tell you this," confides Deborah Buffalin of the Digital Media Center, who pulled this reporter aside. "But some of these students have been waiting two years to take this class, just for this."

Okamura, 34, watched the competition with a wry smile, her elfin features framed by jet-black hair. "I like doing artbots because it forces the left brain and the right brain to work together," she says.

Artbots are an entertaining sideline, but her real work deals with using robots to transmit a more literal kind of feeling--haptics, or the science of touch. Okamura came to Hopkins in 2000 from California to set up the school's Haptic Exploration Laboratory, with the goal of finding new ways for robots and humans to interact through touch. Her work, and that of her students, has a distinctly medical bent. She collaborates with doctors toward advances in surgical robots, hundreds of which are used every day to perform the most delicate surgeries, where a sense of touch could mean the difference between breaking a suture--or worse--and a successful operation.

"Artificial intelligence has not had the promise everyone thought it would," Okamura says. "We thought we'd have autonomous robots that would go and clean our houses and you wouldn't have to interact with them in any way. It's just becoming clearer and clearer--to me, at least--that to make robotic systems practical they need to co-exist with humans and assist humans. This haptic interaction is an important component of that.

"Here's one way to think about robots," she continues. "They are sort of the arms and legs, the action, the drivers in the physical world of information technology. There are computers all around us that help us with lots of different things, but they can't help us in the physical world. That's what robots are--they're the interface between computation and doing things in the real world."

"When I was really little I wanted to be a physicist," Okamura says. "But then I realized that when you do physics, it's either so big it's on a universal scale or it's so small it's atoms and things on a quantum level. . . . Robotics are really on a human scale."

A few days after the artbot competition, Okamura's graduate students present their final projects in her Hopkins classroom. On a high shelf on the side of the room is a joystick-controlled Radio Shack robot claw next to a game of Operation. Replace the claw with an expensive, many-jointed medical robot and the guy with the light-up nose with an actual patient, and you have a pretty good idea what Okamura is working with.

A microwave in the corner of the room bears a sign reading food use only. Someone has added a Post-it note: robots don't like being microwaved.

Around the room, student projects tackle the various challenges of giving robots a sense of touch. At one desk, students Matt Moses and Michael Kutzer show off a device that allows its user to poke a finger into a virtual environment, feeling the resistance of walls and moving a computer-rendered ball around a screen. Another group has attached a copper tube to a robotic arm, allowing the user to feel different strengths of wind as they move a stylus through the air.

"That's a really challenging thing," Okamura says. "How do you provide tactile feedback through tools?" The forces you sense through your skin--texture, temperature, pain--which are important? Which ones are necessary? One of her past students added dots to a computer surgery display that change color depending on how much force is being exerted on a tool. That one has already proved popular with surgeons, she says, and will probably be the first to make it into an actual operating room.

At another station, a group has developed a feedback system for prosthetics, routing a sense of touch to vibrating coils embedded in a shoe insole. One of the team, research assistant Kelvin Liang, says the project was inspired by the Defense Advanced Research Projects Agency (DARPA), which earlier this year gave Hopkins $30 million to create a mechanical arm that returns feeling directly to the user. Okamura serves as a consultant on the project. DARPA's interest is fueled by the number of amputees returning from Iraq. Okamura points out that the technology, once perfected, will quickly trickle down to civilians.

"There are so many parts of the project that are just so fundamentally challenging," she says. "The melding of neuroscience and engineering--and it's going to help so many people."

Outside the classroom door, a rowdy group of undergraduate engineers are demonstrating their final project: shooting a tennis ball into a trash can. It's unclear what the assignment was, but Okamura gets up to close the door. "I heard someone say this one involves flames," she says.

The crowd outside is "depressing," she says, not because of the intensity with which they aim modified potato guns at a wastebasket, but because it contains few female students. As a rare woman roboticist, it's a shortage she worries about.

Earlier this year, Advanced Robotics, the journal of the Robotics Society of Japan, published a special issue on women in robotics, with a dozen articles by prominent women in the field including Okamura. "Just the fact that that kind of thing exists keys you in to the fact that there aren't enough women involved," she says.

Female faculty, she says, tend to attract female students, but it's a "leaky pipeline" as women drop out of the field before reaching higher levels of education. She says it's getting better, particularly in her specialty, where medicine meets engineering.

"The world is starting to understand more and more the human aspect of engineering--it's not just geeks that sit at a computer all day," Okamura says. "It's working with people.

"I think the thing that's made me really interested in this in the long term is it's so interactive. Some [academics] just do their research and they get an equation, or they do their research and they may have a program that runs and does something. I get to play with what I create."

Email Chris Landers