RoCo, the world’s first expressive computer

Inhabiting a back room in the Massachusetts Institute of Technology’s Media Lab, the robotic computer has a monitor for a head and a simple LCD screen for a face. It expresses itself using its double-jointed neck, which is equipped with actuators that shift the monitor up and down, tilt it forward and back and swivel it from side to side, rather like Pixar’s animated lamp. An attached camera can detect when its user moves, allowing RoCo to adjust its posture accordingly.


Among a growing number of researchers who are investigating how far a robot’s physical presence can influence people. Harnessing technology to manipulate someone or shape their mood is nothing new. Researchers at Stanford University in California have shown that virtual-reality characters are more likeable when they mimic our facial expressions (New Scientist, 11 June 2005, p 25), and that an in-car assistance system can make us drive more carefully if the voice matches our mood. But because robots share our physical space, they can have a greater impact. “If it can actually touch you, it is a lot more meaningful,” says Cynthia Breazeal of the Media Lab, who created RoCo with her colleague Rosalind Picard. “Robots can engage us like never before. They can really push our buttons,” she says.


teams has begun to explore how RoCo’s powers of expression might be harnessed. One aspect they are looking into is how the robot might affect our posture while working. Previous studies have shown that someone’s emotional state can dramatically affect their performance on analytical tasks and that posture can play a role in this. For example, one study showed that people who felt depressed were more persistent at carrying out a tracing task if they were able to slouch while they did it, while cheery souls tended to be more tenacious when they sat up straight – a response known as the “stoop to conquer” effect. “Emotion informs cognition, people whose emotion is inhibited don’t perform intelligently,”

However, people don’t necessarily always sit in the “right” position for their mood. Enter RoCo. By taking up different positions, RoCo can encourage people to change their posture too. To investigate whether this would be useful in a working environment, the team gave two groups of people either a simple or unsolvable task to carry out. Once primed with a feeling of success or failure, the volunteers were given a new set of puzzles, projected onto RoCo’s screen. They found that the successful people who sat in front of an upward-tilting RoCo were more persistent in attempting to complete the second, unsolvable puzzle than those given a slouching RoCo. In contrast, those primed for failure by the previous task performed better when given a slouching RoCo than those given a neutral or upward-tilting one.

RoCo could be programmed to adopt the right posture to foster greater attention and persistence in children, or simply to help improve people’s posture when sitting at a desk. In future the group also plans to investigate programming RoCo to mimic the posture and gestures of its users, just as it did with mine, to see if that builds rapport and enhances engagement with the computer, as such behaviour is known to do