12-12-2005, 03:52 PM
Robot friends for autistic kids
Barbara Gengler
The Australian
DECEMBER 13, 2005
A COMPUTER science teacher at Yale University has been constructing humanoid robots that match the size, speed and range of motion of a one-year-old child.
Researchers are planning to use the robot as an interactive diagnostic device with children at risk of autism, a brain disorder that interferes with the ability to communicate and relate to others.
Robotics researcher Brian Scassellati says the long-term goal of the project is to build a humanoid robot about the size of a one-year-old that can handle meaningful social exchanges and acquire behavioral skills from a human instructor who has no training in operating the robot.
"In the past three years we have designed and constructed the majority of the mechanical structure of the upper torso including the head, torso and arms," Scassellati says.
The robot being developed is capable of simple hand-eye co-ordination tasks, such as reaching out to touch an object, perceptual tasks including visually identifying people, and cognitive tasks such as recognising itself in a mirror.
A robot that can interact with people using natural social cues, such as gaze direction, vocal intonation and pointing gestures, would have great practical impact, he says.
It would allow naive users to interact with the robot in a more natural and effortless way, to command the robot through social instructions and to integrate these machines more readily into their daily lives.
Scassellati says there are many technical challenges in building social robots.
Designing and constructing a robot that can produce gestures and utterances that can be easily interpreted by a human observer is a challenging mechanical design problem, he says.
A more difficult technical challenge will be to build machines that can recognise human social cues such as pointing gestures and tone of voice, he says.
Existing research succeeds in recognising a few of these cues in structured situations, requiring visual scenes that have a constant background or audio signals that contain only the voice of a single speaker, Scassellati says.
Source: The Australian
http://australianit.news.com.au/articles...69,00.html
Barbara Gengler
The Australian
DECEMBER 13, 2005
A COMPUTER science teacher at Yale University has been constructing humanoid robots that match the size, speed and range of motion of a one-year-old child.
Researchers are planning to use the robot as an interactive diagnostic device with children at risk of autism, a brain disorder that interferes with the ability to communicate and relate to others.
Robotics researcher Brian Scassellati says the long-term goal of the project is to build a humanoid robot about the size of a one-year-old that can handle meaningful social exchanges and acquire behavioral skills from a human instructor who has no training in operating the robot.
"In the past three years we have designed and constructed the majority of the mechanical structure of the upper torso including the head, torso and arms," Scassellati says.
The robot being developed is capable of simple hand-eye co-ordination tasks, such as reaching out to touch an object, perceptual tasks including visually identifying people, and cognitive tasks such as recognising itself in a mirror.
A robot that can interact with people using natural social cues, such as gaze direction, vocal intonation and pointing gestures, would have great practical impact, he says.
It would allow naive users to interact with the robot in a more natural and effortless way, to command the robot through social instructions and to integrate these machines more readily into their daily lives.
Scassellati says there are many technical challenges in building social robots.
Designing and constructing a robot that can produce gestures and utterances that can be easily interpreted by a human observer is a challenging mechanical design problem, he says.
A more difficult technical challenge will be to build machines that can recognise human social cues such as pointing gestures and tone of voice, he says.
Existing research succeeds in recognising a few of these cues in structured situations, requiring visual scenes that have a constant background or audio signals that contain only the voice of a single speaker, Scassellati says.
Source: The Australian
http://australianit.news.com.au/articles...69,00.html
