MIT's thinking toys


THE alarm rings but the toy man asleep on his green plastic bed does, not stir. He has a system to get him moving in the morning, however, and the bed soon lifts and dumps him onto a conveyor belt that brings him into the kitchen for breakfast.

When he is dressed, he walks next door to the chocolate factory, which is already humming with pulleys and more conveyor "bells.

The alarm clock bed and the Wonka factory are products of children's imaginations. The computer programs that run the toys were created by children at a Lego funded research project at "Toys that Think", in the Massachusetts Institute of Technology's Media Lab.

Other creations include a city of the future with moving pavements, and a house that allows only certain colours of car to park in its garage.

These childhood fantasies are an important part of the project, which aims to create programmable toys that children can use to express themselves. "It doesn't have to be cars or factories," says Mitchel Resnick, who heads the Lego MIT project. "We've created kinetic sculptures with kids using this technology."

The toy industry has long been interested in play objects that perform. Around the turn of the century, jack in the boxes and other mechanical toys were a big success. In the 1960s, battery operated toys were all the rage, as the toy sector worked out how to make dolls cry and miniature cars to move.

Resnick says the difference between these and thinking toys is that in the old toys, behaviours are dictated to children. His new computerised devices, on the other hand, allow children to decide what the toys should do. "It's a much richer learning experience for the children," he says.

His research, which began nearly 10 years ago, has already yielded commercial products for the Lego group. Dacta, the company's educational division, has been selling programmable toys to schools across the US since 1990. Lego will not say how much revenue the toys bring in, but the company does reveal that they account for half the division's sales.

"Children love these products, which are built with light sensors, touch sensors and other things for them to program as they want," says Marsha White head, marketing manager for Dacta. But because the sets require a link up to a PC, and because adults have to help children to write the programs, the toys haven't made it to the shelves of high street toy stores yet.

Using the latest technological advarices, Resnick hopes to change that. The new phase of the MIT project aims to create programmable toys that don't have to be hooked up by wire to a personal computer and will allow even young children to dictate self created programs with no adult supervision.

Miniaturisation and wireless technologies will help make this possible. Resnick has created what he calls the "programmable brick" a device slightly smaller than a television video tape, which can relay relatively complex behaviour patterns to the toy into which it is incorporated. "Most of this is batteries," Resnick explains. "I think we can make it much smaller.

Alternatively, toys could be geared to accept wireless signals from PCs to give children more leeway in the behaviour. "One of the problems Lip to now has been that a toy that's physically hooked to a personal computer has limited possibilities,

Resnick says. "You could create a great monster, for instance, that opens its mouth when you clap, and moves when you shine light on it, but if it can only move a little way, it's not much Fun. We want something that is more mobile, so you can send the monster to the next room, if you like."

Yet Resnick concedes that making the products smaller and more mobile is only part of the challenge. One of the biggest problems in bringing thinking toys into the home is simplifying the interface so that children can have fun with the objects on their own.

Because of the interface problem, Lego's computerised Dacta products are now only offered to older children (aged from 11) and must be used with the supervision of computer literate adults. Children tell the supervisors the sorts of things they want to do, but adults usually type in the programs.

Resnick is trying to solve this problem by creating software to simplify the programming process. One experiment uses coloured squares with simple instructions on them such as "go", "stop", "forward" and "backward". Resnick says that by using a mouse, the squares can be, put together on the computers as if the child were building a tower out of bricks.

Yet the more simple the interface, the more limited the behaviours the child can instill. And even the computer brick building exercise may be beyond the understanding of many children.

"The easiest thing to do would be to make a thinking toy along the lines of today's interactive CDs," says Resnick. Most of today's computer programs for children have built in responses. A child clicks the mouse on a dog on the screen, for instance, and it barks. Complex behaviours could also be built into blocks and mobile toys. A Lego robot, for instance, could be made to do a dance at the click of a mouse or the clap of hands.

Although such toys may be commercially successful, Resnick has doubts about their creative value. "The child is not giving his imagination free rein," said Resnick. "He's just observing what the manufacturers decided would be fun for him to see. If we can't get beyond that, we haven't really moved much further than what we got with simple battery operated devices." Or, for that matter with the old fashioned jack in the box.