At first glance, they seem like an ordinary bunch of kids pretending to be Sneetches, characters from a Dr. Seuss story. They chatter animatedly and spend make-believe money to get stars on their bellies and join the in-crowd, just like in the book. What's not immediately apparent is that the kids are engaging in some high-tech wizardry. The playroom is a research lab, its walls covered with whiteboards, flowcharts, and bits of electronics. When the kids crawl through a special cardboard box, hidden computer modules activate illuminated stars on their costumes and subtract fees from their imaginary bank accounts. In return, the starred Sneetches, but not the unstarred ones, can interact with a fanciful blinking toy.
The Sneetches are actually participating in a serious experiment devised by two professors from the University of Maryland at College Park that may determine the future of computing. Allison Druin, an expert in children's technology, and James Hendler, a computer scientist specializing in artificial intelligence and robotics, were fed up with the cumbersome PC. They wanted to see what they could do by shrinking the electronics and embedding them in everyday objects.
In five years or so such "ubiquitous computing" might lead to conveniences ranging from doorknobs that recognize the person grasping them to shower faucets that remember the favorite temperatures of everyone in a household. But to develop the right technologies for those tasks, Druin and Hendler first needed a focus group of creative, outspoken testers unafraid of unfamiliar gizmos. In short, they needed kids.
Kids tell Druin and Hendler what they like or don't like, unfettered by adult preconceptions about how computers are supposed to work. And they love to play, which is helpful when testing rough prototypes. Youthful imagination can skim over technological gaps, with no need for fancy mockups. That cardboard enclosure represents a personal elevator? Fine. That robot rolling around the room is actually going into a kitchen, opening the fridge, pouring a glass of orange juice, and setting it down on a table? No problem.
The Sneetches game is an example of Druin and Hendler's work in progress. The computer modules in the cardboard tunnel are small, each about the size of a candy bar, but come equipped with sensors to detect light, touch, or the presence of other modules. And the devices are deceptively simple— just a few chips, batteries, and switches held together with masking tape. Under the supervision of a larger computer, connected via another wireless link, they can communicate with one another, turn on lights, buzzers, or bells, or trigger synthesized voices. Druin calls these modules "magic" because of the way they invisibly add intelligent behavior to inanimate objects. When one of the children crawls through the star-making machine, a module in the box signals another module on her belt to switch on the light that denotes her newly elevated status and instructs a nearby computer to display her shrinking pile of play money. Or the rules can be totally different: The kids can reprogram all of the components using a simple graphical interface.
Everything about the experiment is designed to foster creative synergies between researchers and tykes. The floor is littered with cardboard constructions and crayon markings, a deliberate attempt to subvert an overly orderly, adult approach. "We don't want to take away the fun kids have building things," Druin explains. They can attach a magic computer module to a cardboard box, a stick, or a rubber ball, and it becomes an interactive object to play with. The box transforms into a house that can count how many kids are inside it, and an enhanced piece of wood becomes the key to open the door of the house, turn on the lights, and start music playing.
The signature member of Druin and Hendler's menagerie, the Pets2 robot, is designed to act out stories not just with motion but also with emotion. It can move various parts of its body to express a wide range of moods. Right now it spins when it seems to be happy and hangs its head while looking from side to side to appear lonely. It would take only a few minutes to substitute new body language. Such changes take half a dozen clicks of a mouse, but graduate student Jaime Montemayor is putting together an even simpler interface so that the children can program the robot by pantomime. The children also create their own stories for the robot to act out, making it function as a sophisticated puppet. Pets2 offers the hope of a high-tech plaything that can hold a child's interest for years rather than days or hours.
In response to the children's demands, Pets2 has a soft, brightly colored body and interchangeable arms and facial features. Above all, it is huggable. Every time the adult members of the design team tried to go back to hard metal or plastic, the kids set them straight. "It's impossible to remember how you thought as a child," Druin says. She included the junior members of the team in the earliest brainstorming sessions to make sure the group didn't produce brilliant technical answers to the wrong questions. At first the youngsters treated design meetings like camp or school, deferring to the adults. After six months or so, however, "they see that they're taken seriously, that things change because of what they say," Druin observes.
Houman Alborzi, another graduate student in the lab, is responding to kids' requests and developing a better way for the magic pieces to interact. The key is an improved position-sensing system. Each wireless unit will measure the strength of the radio signals it receives from the others, thereby enabling it to calculate how far away they are. Armed with all the relative distances, a central computer will then be able to figure out the overall positions.
The most immediate application of all this child-directed technology will be educational toys that are genuinely fun and helpful for a change. The group has taken Pets2 into local Maryland schools, to an enthusiastic reception. A mass-produced version, with spare arms, eyes, and noses, might reach stores by the end of 2001. Interactive storytelling built around Pets2 could help teach kids about narrative and encourage them to develop tales of their own. The researchers are also adapting the robot to help disabled children exercise and express themselves.
And that's just the beginning. Druin wants to push the starting age for computer-aided learning down to the preschool level. She envisions magic modules embedded in the blocks children play with, as well as in the classroom tables and desks, so that letters light up and speak when strung together to form a word. Or touch-sensitive pads that allow children to "finger paint" on electronic walls, allowing half a dozen of them to paint at the same time, save the resulting mural, and then wipe it clean or turn it into an animated cartoon. Druin's goal is to encourage the messy, exploratory kind of learning that children do before they learn to sit quietly behind a desk.
The lessons from all this kids' play will guide how computers get incorporated into adult applications. Soon you might find yourself making phone calls simply by tapping on a card in your magic Rolodex, or triggering the answering machine to intercept all phone calls when you push your lounge chair to full recline. You might find such intuitive computers embedded into almost everything you own.
Will you like having all these hidden conveniences? You may have no choice. Children who grow up playing in computer-enhanced classrooms will probably have little patience for fixed keyboards and hard, bulky monitors. As they come of age, they will drag the rest of us into the world of hidden computing, whether we want to go there or not.
See the Web pages of Allison Druin (www.umiacs.umd.edu/~allisond) and James Hendler (www.cs.umd.edu/~hendler), or check out the section on PETS (Personal Electronic Teller of Stories), at www.umiacs.umd.edu/ ~allisond/kidteam/robot-index.html. Mark Weiser's ubiquitous-computing site (www.ubiq.com/hypertext/weiser/UbiHome.html) has useful links and information.
