TOKYO: A toddler says 'nighty-night' and covers his silver-coloured friend with a blanket. This is an everyday scene at one US nursery school, where robots have been introduced among children to find out what it takes for machines and humans to develop long-term relationships.
The Sony and University of California (San Diego) joint project is adding evidence to the notion that children, with their open minds, can welcome and even develop emotions toward the robots. The project is led by Machine Perception Laboratory director Javier Movellan. "The children adapt themselves to the robot and empathise with it, although nobody teaches them to do so," says researcher Fumihide Tanaka of Sony Intelligence Dynamics Laboratories who works on the project.
The children, 24 months or older, started spending one hour every day with Sony's Qrio in March last year at a San Diego school. In one experiment, researchers brought in the robot to take part in the children's dance sessions and found that the toddlers would spend longer in the room if the humanoid was among them.
Tanaka says researchers are increasingly convinced that children consider the robot not a toy or a living human being but "something between the two", a difficult idea for adults.
To contrast with the humanoid, researchers also gave the toddlers a simple toy that looked like a robot, which was handled roughly.
Children initially stayed away from the humanoid. It would often fall over due to manhandling by the children. But after one to two months, the children would help the robot back on its feet. Within three months, the toddlers would never allow the robot to fall.
With the humanoid becoming a playmate, another robot, Rubi, which runs on a wheel with a TV panel in its belly, joined the class in April 2005 as a teaching assistant.
The project involving the children could help researchers develop more sophisticated robots. Human brains are "very good at handling uncertainty in everyday life, which is difficult for current robots", says Movellan. "It is important to understand real-time social interaction," he adds.
